Well, here is my latest theory. I reset the learning on my 812Sand went to burn 4 more +RWs in a row. This time I used old discs that had already been burned. The first burn was moderately bad but no where near as bad as beore. By the 4th burn I was back into the teens on my PI scans.
I am guessing that the problem lies in Liteon's learning memory. When I first used these RWs they were all new and unused. These are what Liteon learned and stored in memory. My guess is that once burned RW discs are different from new RWs and they must be burned differently. Conserquently, Liteon does poorly with used RWs when it has learned to burn these from new discs. Since NEC does equally well with either my guess would be that they learn on the fly each time. I could be wrong but it makes sense. C0deKing, let me know what you think.