To me, the changes these categories are describing should be defined by how much they alter the way a CAP performs in the metagame. As such, I'm interpreting major changes as those that would drastically alter the way a CAP performs in the metagame, significant changes to a CAP's established identity, or large changes in how viable a CAP is in the metagame. Mainly, these changes should be done to address a major flaw in a CAP's balance or usage, . Furthermore, I'm using the term minor changes to describe the CAPs I believe need substantial updates to movepool or abilities to remain competitive in their current roles. The main difference to me between major and minor changes is that one is attempting the correct a serious issue in a CAP's design, while the other is concerned concerned solely with keeping the CAP relevant this generation. Simply put, one's addressing design problems which will probably change how a CAP works, while the other is tackling relevancy issues to make sure the CAP continues to work. As a result, consistency for me will be defined as almost janitorial duties we'll be performing to make the CAPs more polished in the current generation. I predict there would be considerable changes done with these Pokemon, but they'd be more along the lines of adding a handful of viable moves from later gens as opposed to complex changes. In short, major updates for the the CAPs that really need major changes to function properly, minor updates for the ones that need the change to remain relevant, and everything else is under consistency.
Now that I've defined what these categories mean to me, I'll sort the CAPs into them along with some reasoning and feedback.
Starting off with major changes, the few CAPs I see potentially falling under this category are Aurumoth, Tomohawk, and Voodoom.
Aurumoth is the least balanced CAP in the current metagame, and presents enough of a problem to the meta that some aspects that define it necessitate a change. Ability and movepool will both need to be looked over thoroughly to find the best way to balance Aurumoth without gutting it or doing too little. Not much else to say here other than Illusion is a god awful ability.
Tomohawk is one of the defining forces of the CAP metagame, and is in large part responsible for how the tier has developed for the last several years. While I don't believe Tomohawk is the unhealthy element many have made it out to be, it's certainly as centralizing as a Pokemon gets in CAP. I won't go into whether or not that centralization is healthy for the meta at large here, so I'll just say this- regardless of what we actually do with Tomohawk, there is going to be an intense discussion about what it should and shouldn't do, and because of that discussion and the possible changes Tomohawk may be seeing I'm labeling it as a major.
Voodoom is the CAP I predict to see the largest number of changes, both in quantity and performance, simply because it is without a doubt the least viable CAP, and has been the least viable at least since XY. I've seen the best teambuilders and best players the CAP metagame has to offer try to make Voodoom work, and none of them have ever been successful. The metagame is just so hostile to Voodoom I find it hard to believe we can salvage it with anything less than major changes. If we make Voodoom halfways usable without changing everything about it I'd call that a win.
Continuing on to minor updates, the big three I see falling under this category are Arghonaut, Revenankh, and Kitsunoh.
Arghonaut currently has no real role in the CAP metagame for a number of complex reasons, but they can all be traced to its horrendous matchup against many of the largest threats in the metagame. I think the update process for it will have to accept that and instead focus on defining what Arghonaut is meant to do. However, unlike Voodoom where the current CAP is simply ass, Arghonaut already has a lot of strengths to work with, and as such should only take a few tweaks here and there to make work.
Revenankh is similiar to Arghonaut, but is much more defined in what it can do and what it's established identity is. Unfortunately, Revenankh is another case of matching up very poorly with top threats, and is again something the update process for it will most likely have to accept. A handful of well thought tweaks to movepool and maybe ability will go a long way towards keeping Revenankh relevant, and we shouldn't have to gut the mon to do so.
Kitsunoh is the odd bird out of the three Pokemon I mentioned, as it's actually a decent Pokemon to use and has a good established identity. However, I believe Kitsunoh is by far the weakest out of the remaining CAPs because it's simply okay at what it does. I've always felt like Kitsunoh has tended to stay afloat by banking on its three best aspects, high speed, great typing, and decent support moves, while the rest of it's package is relatively mediocre. Kitsunoh, in my eyes, will need more of an overhaul than anything else in consistency to remain viable this generation, which is why I've placed it as minor. That said, it is the CAP i'm the least confident in ranking, so I can see good arguments for it being grouped with the rest of the consistency changes.
And with that, everything else is a consistency update. I'm sure many won't agree with, but outside of the above mentioned CAPs, I think the rest of them can easily fit under this category because of one major reason: they already work in the current metagame. If the Pokemon work in the metagame, are viable and have the tools to remain viable, I don't believe there is much for us to do besides adding some generational updates. This is apparent with CAPs such as Colossoil, Cyclohm, Volkraken, Mollux, Crucibelle and its mega, Pyroak, Stratagem, and Syclant, who have consistently been some of the top performing CAPs in the metagame, and have an extremely solid established identity that needs real change. Changing these Pokemon in a significant way is just begging for us to make a mistake and ruin something good we already had.
Now, I've seen quite a few people include some of the more middling CAP mons as minor or major changes, which is a stance I wholeheartedly disagree with. There is nothing wrong with Plasmanta, Krillowatt, Naviathan, Necturna, and Fidgit, and I really do mean it when I say this. Simply put, these Pokemon perform well enough in the current metagame that changing them is unneeded due to the number of strengths they have. They may not be the greatest performing or the most viable, but any perceived lack of ability is because of a hostile metagame or a niche that will never be destined to be dominating. Plasmanta is an incredibly powerful Pokemon thats held back mainly by its middling speed and bulky Ground-types, specifically Colossoil. If at any point the metagame slows and AV Colossoil is not suffocating the tier in usage, I can assure you that it'll be a major threat. Krillowatt is very similiar to Plasmanta, but in my opinion more limited than it because its entire movepool is unfocused. Arguments on removing Magic Guard seem to be based on the fact that LO Magic Guard has been its defining set since creation, but that's not because Magic Guard outclasses Trace; it's because Trace is just a shitty set that has nothing supporting it in Krillowatt's movepool. Krillowatt really only needs two moves from generational updates to remain viable, maybe even a lot more viable, but overall I think Krillowatt can skate by with a very minimal amount of changes. Naviathan is being discussed because Dragon Dance is an underwhelming set and CM has trouble breaking stall when it's Dugtrio bait. That said, Naviathan's Calm Mind set would be a perfect stallbreaker if it wasn't weak to Dugtrio, and it's Dragon Dance set suffers mainly from damage output issues, which can be resolved with very minor changes to movepool. Again, in a less hostile meta Naviathan would be a very good metagame pick, so depending on how far you want to go with pushing Dragon Dance it's either a minor or consistency update.
Necturna is the biggest can of worms we could possibly open. I could not recommend any changes that make Necturna less reliant on Sketch this early in the generation; unlikely every other CAP Necturna has an insane amount of potential sets we've yet to see. Also, Necturna has a very strong niche already as the hands down best Sticky Web setter, so viability and relevance is hardly and issue for it. There's many other reasons I have for not wanting Necturna getting more tools, but I'll save it for the update process for it. Fidgit is another suggestion I don't understand. I guess people are assuming because theres three gen's worth of utility moves we're going to be loading them into Fidgit like an overstuffed taco, which I honestly hope we don't do lol. Fidgit already defines twoplay styles in CAP, and has proven over the last year or so that it can perform in roles indepenent of setting a field condition, so there really is no need to change that. For now I'll be considering Fidgit a consistency update until I see some actual evidence that it needs a buff of any sort.
Another Pokemon I'd like to address specifically is Cawmodore. Echoing what reachzero said, there is no need to change it. There is nothing, and I mean nothing, that points towards it being unbalanced in anyway, nor is there anything that points towards it being an unhealthy aspect of the tier. I don't buy a single one of the arguments that center on Cawmodore being an instant lose if it sets up, an instant loss if you don't pack a dedicated check, or that it simply sets up on too much. To me, none of those argument have any evidence to support them in the current metagame, because each on of them would result in a top tier Pokemon if they were true, and Cawmodore is far from being a top tier Pokemon. It has a host of weaknesses and pitfalls that are more apparant than they were last gen, which really demonstrate how flawed it would be to change it.
The last CAP I want to talk about is Malaconda; while I don't disagree with the claim that it is a very lackluster Pokemon in the current metagame, I have to point out that Malaconda is one of the most specially bulky Pokemon in the tier, with a serviceable defensive typing and a host of other strengths that really lend itself towards trapping and removing a handful of Pokemon. Unlike the other underwhelming CAPs I've listed as minor, Malaconda is competent in the small niche it fulfills in the current metagame, and has the potential to become significantly more viable independent of what we do.
So uh, tl;dr: major changes for balancing, minor for the mons that need a bump in usability, consistence for everything else thats perfectly usable in the meta. I'm not a fan of changing mons for the sake of changing mons, and I won't advocate a major/minor change unless there's solid reasoning behind it. Reasoning > all in my book.