Pointed my miners at theblocksfactory. Initially I was disappointed, effective hash rate for myr-gr with ccminer v8.11-klausT was the same 25% of the miner reported hashrate I had seen on suprnova.

After changing the bat file to use ccminer v2.1-tribus instead of v8.11-klausT, I'm seeing an average effective hash rate that matches the miner reported hash rate, yes!

I'll be back to dgbg.suprnova.cc if/when it and some version of ccminer myr-gr interact as expected.

Also, something appears to be wrong with the myr-gr algorithm in ccminer v8.11-klausT. If I let theblocksfactory vardiff do it's thing, sgminer, reporting 115MH/s, switches between diffs of .128 and .256. v2.1, reporting 152MH/s, switches between diffs of .128, .256, and .512. v8.11-klausT, running on the same two cards as v2.1 and reporting 163MH/s, switches between diffs of .032 and .064. Even at this lower diff, v8.11 submits many fewer shares in the same time and has an average effective hash rate of 25% of what is reported by the miner. This probably isn't a pool problem as it happens on more than one pool.  

Thank you for recommending theblocksfactory.