DGW3 still adjusting? We are at 1760 new blocks now, 1760 calculations to adjust. I can't see any reason in the code why it could take that long.
Because of the spike at the start from 0 to 9 GH/s. It was adjusting to that and meanwhile the multipool added swings as well between 5 GH/s and 11 GH/s. Couldn't handle that all from the start? Too much to handle at once. But maybe I am wrong about this.
Fact is that the big gaps in blockfindingtime are not there at the moment.
http://www.guldencointrader.nl/When the big gaps come back, then we know for sure that this DGW3 isn't the best solution and needs some adjustment or replaced by something else.