I'd like to thank everyone that's participated. There's been great discussion on the initial stat limits and some great argument in favour of changes. It's been difficult to narrow down everyone's perspectives since there are a lot of differing opinions about the direction to take CAP 28 in but here are your final limits.
Code:
PS: 220
SS: 225
PT: 120
ST: 180
Minimum PS or SS Mask: 140
Total BSR: 350
Maximum Speed Stat: 144
I'd like to summarise the changes that have been made as well as my explanations for them.
The first change was the most intuitive and that was the increase of the PS and SS Mask. 120 undershot where I was hoping, and allowed for purely Physical or Special sets to maximise their tankiness as well as single sweepiness without significant penalties. 140 as a mask allows for pure Physical and Special sets to subsequently choose between hitting their maximum offensive or defensive benchmarks, while still allowing for investment in both.
Next Both PS and SS have seen an increase due to effective arguments that they slightly undershot some key offensive benchmarks we may want to hit against some of our most important pivots, and subsequently incentivised running more bulk, which goes slightly against where our previous discussion went. As many have pointed out, although having equal PS and SS limits seems intuitive, Special spreads are not actually able to reach the same benchmarks as Physical spreads due to the calculations. Factoring in the fact that most special moves are also slightly weaker overall, adding an additional 5 to the SS limit allows special sets to reach the appropriate benchmarks.
I decided against raising the PT and the BSR limits even though there were some arguments for them, because I believe that doing so would once again encourage pure Physical or Special spreads to hit the other stat limits without trade offs. While maintaining the BSR of 350 will still affect mixed spreads more than the purely Physical or Specially spreads, I believe that raising the BSR limit will not actually solve this problem, but allow pure Physical or Special spreads to become even more efficient.
As for our maximum Speed limit, I've set it at 144 because higher stats will not change the Sweepiness benchmarks. While I initially decided upon a lower limit, because it was something that I didn't specifically raise in discussion questions I don't think that it would be fair to throw in a number that might disrupt some preliminary spreads.