Okay, so why not put EVERYONE into a protected class of some sort?
(oh, and BTW, unions are awesome!)
You have to understand that there is no correct universal objective solution to this. There are only different versions of pros and cons and all are diferent versions of subjective good and bad.
If you use one set of axioms you get one set of results for good and bad. If you use another set, you get a different result. If you then treat your set as the universally correct one, you are doing something which is not true.
Rather if you really want one of the core axioms, that is a limited one, that you can include, here it is.
In some cases good and bad is individually and bad can happen to another human, if someone treat good and bad as based on their indidual understanding.
In other words for your method, you can accept that bad can happen to somebody, because it is overall good. In my version good and bad are different, because it is not done the same way for overall good, because I use another set of axioms.
That is it. But as long as you in effect argue as if your version is uinversally correct, then we can't explore other versions, because you always default back to your subjective ideal of objective as you do it.
Regards
BTW - everybody is in a proceted class in some systems, because we have minority rights. But you in effect argue for majority rights.