For almost all niches, I would guess it's insignificant. However, I did notice that Keyword Discovery (paid version) recently switched their default database from the global database to the premium, because they felt the global had a lot of "search engine noise" from stuff like automated queries.
After the switch, I noticed that almost all the keywords I work with dropped in estimated volume (obviously, since there are fewer searches in the premium database), however, the changes definitely weren't proportional for all niches. When I started checking out volumes in each database on stuff like seo, viagra, forex, and ringtones, the difference was enormous.
So yes, I think that human or automated queries to check rankings can make a big difference on certain things, but is Google filtering those out somehow? Hard to say...maybe if someone took the time to check the premium Keyword Discovery database against terms in Google to get an idea of whether the predicted volumes are proportional, you could see if the viagra/seo/forex type stuff is similarly proportional or likely to be tainted by rank-checking queries - but I don't know how accurate that would be, or if it would even really matter that much.
