From the mda471 post few up, I'm taking it that Harvard have overcome the Ms Piggy constipation, so if you have CEP2 upload sitting [e.g. if device was set off-line to wait it out], you might want to clear them first... might get you below the 2xBOINC cores again to make your host eligible for download again.
It won't help returning the Harvard-part of a CEP2-result if the WCG-part of the same result is still stuck waiting to upload.
Really? It's the _4 that goes to Harvard and -That- is the monster... the last I had was 72MB, sooooo, if this part can be put out of the way whilst WCG has the outage it will be a whole lot faster to get back to a <2xcores download permission state. The _0 to -3 bits are peanuts by comparison, wherever they go, Harvard or Toronto.