All three polls show very similar results:
Charlie Crist 57% Tom Gallagher 29% Undecided 14%
Jim Davis 40% Rod Smith 35% Undecided 25%
Charlie Crist 55% Tom Gallagher 24% Undecided 21%
Jim Davis 29% Rod Smith 14% Undecided 57%
Charlie Crist 55% Tom Gallagher 32% Undecided 13%
Jim Davis 47% Rod Smith 19% Undecided 33%
The consistency of the Republican numbers is actually key. To be sure, I expect the race to be a heck of a lot closer than these polls — but all three show Crist with a lead greater than the number of undecided voters. That can not be good for the Gallagher campaign. Heck, Crist is winning by at least 15% even if one assumes the margin of error for each poll was entirely in his favor.
As a political scientist, I can say these surveys can be fairly accurate. NPR has a good interview from September 2004 with Frank Newport, head of the Gallup Organization, on the accuracy of political polls. It also talks about how pollster determine a “likely voter” using questions, that are fairly good predictors for likeliness of voting.
In the end, it would seem that we will be watching Charlie Crist vs Jim Davis this fall.