Anything besides a signed message is most likely BS.
$ gpg --export 5EB7CB21 | gpg --list-packets - | grep pref-hash
hashed subpkt 21 len 5 (pref-hash-algos: 8 2 9 10 11)
Compare to the well known key:
$ gpg --export 5EC948A1 | gpg --list-packets - | grep pref-hash
hashed subpkt 21 len 3 (pref-hash-algos: 2 8 3)
The 8,2,9,10,11 list was added to the GNUPG code tree in commit e50cac1d848d332c4dbf49d5f705d3cbbf074ba1 on July 9th, 2009, and not released until version 2.0.13 later. This is well after the 2008 date on the key. The 2,8,3 list was the prior list the would have been used in 2008. That they were different at all was surprising, considering that they claim to be generated less than a day apart.
This key was also not on the keyservers in 2011 according to my logs; which doesn't prove it was backdated, but there is basically no evidence that it wasn't and significant evidence that it was. And it's not turning up in any of the older key server dumps.