Online references appear to show no significant or abrupt change in drag coefficient as a sphere goes subsonic.

http://arc.id.au/CannonballDrag.html

Don't be afraid to look at the NASA links, they're pretty good.

The biggest change occurs at about .8 Mach.

Sonic velocity is 1115 fps at sea level in dry air.

There is a well known drag reduction at about .3 M, depending on Reynolds number and a nifty effect on smooth vs. rough spheres that accounts for dimples on a golf ball.

Other than that, I can find no drag spike even at transonic speeds.

I just don't believe the huge variation in actual muzzle velocity vs. 3 foot measured velocity is at all real.

I'm open to actual data. My experiments are admittedly limited, and there may be effects I don't know of. But for now, I'm unconvinced.

"The price of good shotgunnery is constant practice" - Fred Kimble