Ryan thank you for your service!
You’ve asked a broad question here, and a little more specificity will lead to more focused answers.
The big branches of the decision tree are budget, naturally aspirated or turbo charged/normalized and, to a lesser degree, avionics. There a four generations of SR22s, with some variances within each. The G1 airplanes are fantastic values to buy, the fastest of the NAs and offer the highest useful load if that’s an issue for you. Early examples are 6 pack planes, and with your experience you’ll feel right at home. Many G1s have Avidyne PFDs, dual Garmin 430s and a decent enough AP. Buying one of these will require a very thorough review of what will likely be thick MX logs.
The G2 got a arguably better fuselage, as it is constructed with precision molds and resulted in better shaping and less variance. Late G2s saw the intro of the aftermarket turbo normalizing option, and the logs will reflect some teething pains.
The G3 added precision molded wings, taller gear, and the greatest number of airframe improvements along with the brilliant Garmin G1000 panel (Perspective). Probably the sweet spot in Cirrus production.
When time came for a G4, the Chinese had bought Cirrus, and 4 is an unlucky number so they jumped to G5. 200 pound max weight increase thanks to a bigger chute, better flap utility and 3 seats across in the back are the big changes. Turbos have moved away from the normalized aftermarket units to factory true turbo installation.
So there’s your very broad overview. Generally, buying the nicest example you can find will bring the most pleasure. Like all man made machines, some Cirri are reliable and relatively trouble free, while others seem jinxed. Looking at your resume I’d say your big challenge will be the amount of owner involvement required in MX. There are a lot of choices to make, and you will hear many opinions. Many use a third party to advise on MX like Savvy; just remember they earn their fees by managing costs largely, and that may not be your MX philosophy. The big complicating factor is the often poor quality of MX work out there that brings an inordinate amount of MX induced failure to the process that might otherwise seem clear.