I can see why individual users want this, and I can see how it makes their lives a little easier, but I am less convinced of the benefit to the project.
Is there a backlog of identified yet unreverted vandalism needing rolled back? On the contrary, we've edit conflicts of users trying to revert half if it, and the bots usually win.
Now, I guess we can say that there may be marginal benefit in granting it - if it encourages the vandalfighters (do they need encouraging?) fair enough...but....
1) The wastes of time in creating a new "request process"
2) It will lead to at least some disputes, ANI traffic, wheel wars and arbitrations. If I think user x should get it, and you don't....so?
3) If user y thinks I removed it unjustly, where do they appeal? [[Wikipedia:Requests for rollback review]] - I predict, not in jest.
4) It occupies developers' time: have they not more pressing problems to solve?
5) Instruction creep. Endless discussion over adjustments.
6) Abuse of the tool - and the drama that creates
So, on a cost/benefit analysis, this makes very little sense to me.
Doc