Ultimately people go to prison for different reasons, and so generalising as "ex felon" can cloud the issue.
If you see prisons as being full of child molesters, murderers and rapists, then that naturally leads to some conclusions.
If you see prisons as overflowing with the results of the "war on drugs" then there are possibly different conclusions. Are these people criminals" or POWs?
Ultimately every case is different and has its own merits. A teenager can be convicted of no more than possesing some weed. Or shooting up a school. Clearly there could be easier moral paths to reintegration for one than the other.
Even "sex crimes" happen on a scale. 2 consenting 17 year olds, having sex in an inappropriate place, or sexting each other nudes, leads to the same "status" as afore-mentioned rapists. (obviously sentence is different but they'll forever be on a sex offence registry.)
So talking in sweeping terms about inmate populations and reforms, can be counter productive. It may be more helpful to discuss narrower groups, and the nature of the crimes that got them there.
By the way, this is one reason eliminating background checks backfires. If people can’t look up your specific offense they assume the worst. Further, the evidence is that eliminating background checks results in a drop in employment for black men. If people can’t check whether a candidate is an offender, then they resort to broader statistics and patterns for their decision making, and the statistics being what they are, this mostly just hurts black men who have never committed a crime.
See, for example: https://www.journals.uchicago.edu/doi/full/10.1086/705880#d3...
And yes there will be some price to pay.
And as I mentioned in a previous post, eliminating background checks backfires anyway, because people simply stereotype more when you take away their ability to look at specifics.
If you care about reintegrating criminals, then be the change you want to see in the world, but don’t try to tell me there’s anything wrong with someone who chooses differently.