if there's some new theory you've got, that's a different point. excluding things like elastic effect/cross-training/new theories that can make training effects non-linear, coco is right... training fits squarely into the linear systems analysis world of math/engineering.
I stand corrected. (hopefully the only victory coco gets out of this process

)
which strategy would you recommend to managers who draft potentially valuable trainees? #1 obviously hurts the u21 to some degree, so do you have separate instructions for NT prospects vs u21 ones, as magiker alluded to? or do you think one of these is unequivocally better?
It depends...
If the person is training a specialty player or a player that is going to need more secondary training such as a LI PG or a PF or maybe even a R&G Center I would say to go with option 1...
however, as i stated before, specialty players aren't going to win us anything if we don't have them mixed in with what i call the meat and potatoes type NT players... so i would tell the majority of managers to train under option 2.