I guess I see the source of your confusion.  Let me try to clarify.  The main issue became that rewarding further back in the share log so far beyond the block withholding attack shares that were already rewarded naturally would have rewarded miners in a very disproportional manner.  It would have rewarded shares of miners who a) had abandoned the pool, and/or b) weren't affected by the withholding attack at all in the first place, and/or c) wouldn't have been rewarded a sufficient amount to be paid over 2 TBC (~0.001 BTC).  So, there was no perfect solution to this whole thing.  It didn't make sense to make many thousands of dust payouts to miners who had long stopped mining and/or didn't have anything to do with mining during the withholding attack.  At the same time, it would have been wrong to selectively choose miners who were affected by the attack and disproportionately reward shares that would have never been rewarded otherwise.  It would have been wrong to use the massive amounts of < 2 TBC balances as an excuse to suck up those funds as pool donations, too, so that was out of the question.  Rewarding and payout of the shares as was done back in January was the most fair and sensible solution available.  Again, not perfect, because it did not eliminate the issue of rewarding miners who weren't affected by the withholding attack in the first place... but that was an unsolvable problem to begin with.  At least the rest of the issues were eliminated by the final solution.

I was also not going to take any action regarding the coins that were recovered back in 2014 until after a full review by legal counsel to cover both myself and Eligius on all possible risks.  This was not an inexpensive task, either, and Eligius has virtually no real income aside from donations which haven't even covered hosting costs for a while.  I didn't think it right to take away from these held funds due to the miners to cover those costs, either, as they still were not Eligius's funds to spend.  In the end it was an issue that had to finally be resolved, and I took the steps and paid the needed costs associated with doing so myself.  Suffice it to say, not even counting my own time, over the past year or so I've Eligius's biggest donor.

I do believe, in your personal case, you're mixing natural pool luck in with the block withholding attack offset.  I don't have the exact numbers in front of me, but the block withholding attack accounted for a few % drop in shares rewarded throughout the affected period.  There were other periods of bad luck that were natural and unrelated.  The CPPSRB share log was run through simulations related to the block withholding attack period and its expected effect on the share log throughout that time frame, and that data was analyzed extensively before coming to the final reward distribution method.