As far as I can tell, the problem with this approach is that the people who stopped mining may never get paid, at least in the hypothetical scenario you originally laid out as the reason for this system over SMPPS.   The oldest shares only get paid when the pool gets lucky enough to "catch up".  But in that case, this system wouldn't have been needed anyway.

In this scenario, you'll just have pool hopping again.  As soon as any unlucky streak starts, I ought to hop out because I may never get paid for the work I am doing now as these will be the "oldest unpaid shares".  I'm better off finding some other pool that isn't having bad luck and coming back to this one only when it gets lucky again.  This is bad for a pool for the same reason that classic pool hopping is a problem--because if everyone does it. the pool hashrate goes to 0 and the pool dies.

I think what this boils down to is that there is no system in which a small pool with a bad luck streak is a good thing for participants.  In SMPPS, you have the situation you described.  Your RSMPPS has the "abandon an unlucky pool" problem I described.  In your alternative, you proposed, you have the In proportional, you'll have pool hoppers.  In score based, "cheat proof" systems, miners still run the risk of not recovering from the unlucky streak prior to the next difficulty change.

The purpose of pools was to insulate individual miners from the variability of being a small solo miner.  With the difficulty as high as it is, small pools just can't do that no matter what payment scheme they use.