Goldbach is just sending out useless UpperCase WUs en masse, to keep such a destructive project in the stats is imho a bad sign.

It's OK for projects to do some initial testing "in the wild" with UpperCase, but after a few thousand, at most for about 1000 credits, it's just wilful stealing from other projects and it's obvious that the project is just a scam.
Even BelgianBeer needed only ~650 credits to get it all running smooth and rightfully stopped than because nothing useful was developed. Goldbach has so far crunched nearly 600.000 of the useless stuff, 1000x of what was needed by the total amateurs of BelgianBeer.

IMHO a stats page that includes such scam-projects isn't a good stats page.
You can argue whether the science behind some projects is something worthwhile (imho prime search isn't worth a single CPU-second), but at least they are doing science, Goldbach is doing nothing!