Could you please elaborate on the team training results you've been experiencing?
Some solid data would make your claim more credible, or at least we can help interpret and conclude on the strange occurrences.
My understanding is that there is a solid formula used for training, which is then modified by variables and accelerators. Factors such as age, height, existing skill set and sub levels.
With these modifications, they are then applied to the skill. Although I do find it unlikely a skill would experience 1.0+ in pops. You'd have to be on 5.99 to go to 7 with everything working in your favour.