Why not? Some companies do try this. It rarely succeeds though. The reason is that re-education en-masse, restructuring of the chain of command, re-allocation of resources are hard. Most businesses at some point enter the phase where the inertia is the strongest driving force: they only need to apply a tiny fraction of initial force to keep the lights on. At this point, trying to restructure is bound to be very hard.

Maybe, if the company can foresee and realistically assess the problems it's about to face, it may gradually prepare for the transition. I've seen this happen at my friend's job at Dell storage division: some storage product failed, they tried to reshuffle the teams to start working on something else, with some code reuse from the previous one. It still didn't go well, and a lot of people were still let go (because the initial effort of developing a new product cannot really accommodate an army of various kinds of extra personal that's necessary for mature product). They sort-of survived, but with a huge loss.