I think the key part of theorymoning is to develop a stall team that can cover all your key bases. Unlike other teams which can sometimes afford to lose one or two Pokemon, a stall team needs outs and changes that it can use. Is there a mathematical solution where it is inherently possible to cover all situations? Probably not. Deriving a team which covers every and all possibilities seems a bit extreme. But I suspect the emphasis on playing the theory card initially is to ensure that a key threat won't punch 2+ KOs every game.Theorymoning counters rarely work in practice, and application is where it counts, and I also feel that lacking a failsafe (can be a win condition, but mainly just a back up plan) can be detrimental to the success of a team.
That's where practice (and as I put it, putting mileage on your team) comes into play. There's legitimately no way in my mind that you will have accounted for everything and anything against your team. The next step is to figure out what unknowns you missed by practice. Theorymon relies heavily on assuming something in an inherently semi-structured environment. This fine tuning will come through practice. Maybe (for example) Venusaur isn't the best option for your team. Maybe [xyz] comes up more often than not and your coverage of Pokemon like [xyz] isn't reliable because of it. Or maybe combination [abc] and [xyz] is more common. There are difficult to account for individually and in the massive pile of users nearly impossible to solve. It's not not impossible...but there's far too many permutations for any one person to really consider. It's the same way Chess isn't mathematically impossible to solve but there are so many options that research instead has been on developing strategies for the game that is forming in front of them.
Actually, I quite enjoy the comparison of Stall to quantification of Chess. You go in packing stratagem and set plays. Once you're there you determine on the fly which ones will work, which ones you prepared for naught, and adjust your movements accordingly. You play cool and patient; pick off the targets as you manipulate the field and force them into no win situations by forcing key sacrifices.
I (personally) don't frequent the concept of win conditions for my own teams. The design typically runs down to a mathematical "you can't get around [x]" and a mutual agreement of forfeit or we run down a few more turns as I nickle and dime their HP to 0 by Burn, Leech Seed, Toxic, and damage from decent attacks...Heatran Lava Plumes for example). The ability to shift gears on a dime sometimes eats up way more room than I'm comfortable giving...though some Pokemon (Clefable, Suicune) do have that nice ability though. Personally I'm tempted to try a stall team with that ability once Mega Latias comes out.