In docile trainer aircraft the spin can be recovered from, but that's usually only allowed to be tested in the utility category which has more restrictive weight & balance limits. In famously efficient airframes, it can take several thousand feet. In this article, the famously hard to insure pilot killer the Lancair IV took 4,000 feet:
https://www.kitplanes.com/taming-the-lancair-iv/Part of the Part 23 (certified aircraft rules) process is demonstrating spin recovery. The Lancair is experimental so it's not Part 23. Cirrus said, "We put in a parachute instead of demonstrating spin recovery." To which the FAA said, "OK, that works for us."
Since the Cirrus prohibits spins (most certified aircraft do) and Cirrus never demonstrated it officially, I only have the word of random comments on the internet to go with. The assumed state is, "The SR22 has poor recovery characteristics from spins and the manufacturer states in the POH that CAPS is the only recovery method." Maybe it's not all that bad, but at pattern altitude (this has sadly been demonstrated) I'm sure it's fatal.