With the utmost respect for your expertise and knowledge, I have to ask.. Are you positive this is not a glitch in the system somewhere? The mathematical improbabilities of this happening are getting very large.
As I posted earlier, there were two "glitches" in this last week. One was affecting many pools: There has been a spam/dust "DoS" attack on the network that has been drastically slowing down bitcoind nodes. BTC Guild does not currenlty use "empty" templates as a base when a new block is found, because the idea of making blocks with 0 transactions is just wrong. However, as a result, this last week many of the pools were stuck for 4-6 seconds after each new block. Many updates were being done to try to fix this, and recently sipa released a patch to drastically speed up creating new proper blocks.
The other glitch was one server [no others were found to be doing this] was crashing when submitting blocks. I have not yet found out why it was doing this, but I rolled the bitcoind software back and it has not happened since. That would not be a huge impact though, but it would contribute to the last few days of luck (5-10% maybe).
Very little has changed on the server side recently to cause any kind of software glitch (aside from the server that was updated to 0.8.2). I'm actually starting to wonder if perhaps the problem lies outside the pool, possibly in the Avalon implementation of cgminer. The luck has been trending negative since the time Batch 1 started showing up in large amounts. I'm fairly certain BTC Guild has the highest % of Avalon users as a portion of the overall pool speed as well.