Could be a bad idea, in circumstances such as da-brain points out.
Here's my view on Google: IF there was an advantage THEN by the time it is being discussed openly in forums G is working on a solution - which could included filters, penalties, discount points, you name it.
If you believe what you read and what you are told ranking is a function of 100+/- variables, but common consensus is that inbound links from authority websites is the 350 pound gorilla. (Not quite as large as the 800 pound one, but powerful nonetheless.)
The bottom line truth to questions such as this is that in order to understand the role of the variable you need to run tests using multiple (100s to 1000s+) URLs on many IP addresses where - in each "test case" - only 1 variable is altered and studied for effect. People running such tests are not dishing out their findings in public forums, until such time as the gig is up - and then the lemmings rush in to their slaughter.
Run some tests if you have the time, money, servers, . .
OR, come to a PubCon, ply Dave_N with quality liquor and listen very closely. There's some very small chance that you might pick up something worthwhile. :)