I disagree with everything you said here.
Gyroscopic stability matters from the instant bullets are muzzle released, from bullets snapping to sleep to bullets immediately tumbling, and certainly w/regard to accuracy. Why do you think a bad crown hurts accuracy?
In contrast, Sg in no way affects transonic issues (which are matters of dynamic stability, Sd). A bullet can be at a relative Sg of 7 by transonic, and if it's dynamically unstable, it will still tumble there. The infamous Sierra 168gr MK perfectly demonstrates this:
Released at 2800fps, 12tw(12" per turn), std. ICAO conditions, Sg=1.47
By 1100yds the bullet's effective twist rate has dropped to 5.4"per turn, relative Sg=7.17, and velocity at 1029fps (entered transonic)
The bullet consistently tumbles..
You did qualify your notions with <1000yds, and most bullets do not have an Sd issue, but being 'fine' with marginal initial stability is in fact -less than optimum. The reason for this is less than full gyroscopic forces to overcome the huge disturbance that is muzzle release.
Things have happened with bullet flight even before they've made it through your chrono screens.