it wouldn't explain anything. MOST teams change signs frequently in-season. Only the stupid ones don't...
You keep saying this, I'm going to address it.
First, whether or not teams change signals is not related to whether or not Michigan cheated. Barring that multiple news organizations are flat out lying, Michigan cheated. Full stop.
Second, can we please think critically about this. I've seen comments, for example, that a team changed their signals at halftime.
There simply is no way that teams are devising and implementing completely new signals systems at halftime. Realistically, they probably aren't even doing that in a week.
Remember here that every guy on the team has to learn the new system. You don't merely have to train your smartest guy nor even your average guy. A convoy is only as fast as it's slowest ship and similarly your signals system can only be as complex and quickly changed as the slowest guy on the team can keep up with.
Given the limitations outlined above, my guess is that most teams, as a practical matter, utilize a finite number of signals systems. Typically teams have three guys signaling with one live and two dummies. A logical match would be to have three signals systems, one for each signaler.
One of the advantages of cheating by conducting prohibited advance in-person scouting and utilizing prohibited electronic devices is that Michigan may well have been able to determine the number of systems and decipher all of them in advance.