The new honey monkey project which is getting some publicity at the moment has severe limitations, if not doomed to failure.
The whole project is limited in scope, it only uses Windows XP systems (O/S specific exploits) and the MS operators are going to pick the sites they go to!
Like that's gonna work ... One of the prime vectors for these exploits involves phishers pointing their e-mail scams to private sites, not site which would be discovered by exploring links.

The only threat to the phishers from this project is that MS might develop a fix to block their latest exploit (a slow response, which they can develop a response to), or start legal proceedings (an even slower process, which will not be able to touch these fast movers in the dodgy parts of the 'net).

The whole thing is more a PR exercise than anything useful!