An inchworm is a small animal that crawls by pulling its back part of body towards his front part to then extend his front one forward.
Such movement pattern can be used as a mental model to think about progress in all sorts of suits in life.

Improving at fundamentals (i.e. lower end of the worm) serves a solid basis upon which the most advanced skills need to be built.
Unfortunately, the sole increase in one’s skill ceiling does translate directly into improvement for two reasons more or less interwinded.
First, assuming a gaussian distribution, the mean performance is the most impactful aspect long term as it is what is mostly likely to take place and as such is the one that should be optimized against.
Second, the benefits of a strong A game are nullified by the worst games taking place an equal amount of times (due to symmetric probabilities a gaussian curve).

This theory looks to me as a specific instance of the inversion principle, the one by which it can be easier to solve the complementary version of a problem instead of canonical one.