Interesting - I checked to see if I had anything similar and found one pending result - also from May 15 - that has not yet been validated even though both results are back ... perhaps there is something about some of the results returned that particular day ?

CMD2_ 0001-GPDAA.clustersOccur-MYH6.clustersOccur_ 1559_ 1-- 611 Pending Validation 12/05/09 17:11:13 12/05/09 19:10:36 1.08 22.4 / 0.0
CMD2_ 0001-GPDAA.clustersOccur-MYH6.clustersOccur_ 1559_ 0-- 611 Pending Validation 12/05/09 17:04:35 15/05/09 02:02:22 1.48 18.4 / 0.0

The complete quorum/not validating after many hours / days has been mentioned by the tech and reported by a number of users so you can hate yourself astrolab. A flag that should get the validator to muster the complete quorum is not being set sporadically (what's sporadic on 420,000 tasks returned in a day). Periodically, round about when the batch needs completing, moving through the assimilator and sending to the scientists, knreed will do something to force it off, but he is looking at the problem.

Seems to be "where's the FAQ" time today, 99.9% of the time a waste to generate since asking is easier. I'm a volunteer, we all are volunteers, BOINC is open source at that, so, as with the member who was going to collect all the GPU/CUDA info and create a wiki page, you are too welcome to make one. I contributed FAQ's even before being called to the CA ranks. Can't remember having seen this since about 2005.

And all the commenting on large numbers of PV for this project. You know why that is, in amongst the short duration and some hoarders of large stockpiles in their cache so they can be sure before the start-up supply dries up. I still got HCMD2 tasks in PV from 13/16th (14), and no wingman showing yet, that I can assert you. It is AFAICD by no means a wide spread issue... knreed would know for he has detail by-project reports on what's in PV.

Perhaps your queue isn't backed up, but the one with the other compare workunit to be run probably is. My oldest PV is from May 16. Everybody running the project has got 'em. I just keep crunching. If it gets close to deadline, the system will send out a high priority workunit -2 for someone new to run to compare with either -0 or -1 already sitting here.

I can report that my IP units are now current and I am down from the all time high of 146 pages of IP to a measly 12 pages. My one computer that got loaded with the under-estimated jobs has caught up, even though I did decide to abort about 2.5 days of WU as I was afraid that I would not catch up.
While I did have a huge pool of IP, I never had a buffer on any machine larger than 3 days and that wasn't on all of my profiles, and all my profiles are now down to .2 days.