There is no verifiable evidence that Bush carried the Mulsim vote by anything like 70% (other than Norquist asserting it (w/o footnotes)in the American Spectator and elsewhere. It turns out, by the way, that the only assertion one can find after exhaustive research for this claim came from a "sample" done by Sami Al Arian's Tampa Bay Islamic Center. What a shocker! What IS verifiable is that Bush and the GOP lost in every major state with substantial Arab and or Muslim population concentration (CA, NY., NJ, MI -- the latter being the then-only Arab American Senator, Spencer Abraham, in the state with the largest Arab and Muslim concentration). Florida.....well, we all know .... Hope this answers your question.That is some indirect evidence that the Muslim vote did not go as has been presented by Norquist and some others. But there is more evidence, some direct, some indirect.
First are the 2000 exit polls. In the breakdown by religion, 94% of people said they were Protestant, Catholic, Jewish, or "None". That leaves 6% "other" which would be primarily Muslim. The "Other" vote went 62%-28% for Gore. There were some problems with the exit polling in 2000, but not large enough problems to cause a complete reversal within a demographic subgroup.
And the additional indirect evidence comes in the fact that supposedly Bush has alienated the Muslim vote which supposedly supported him so overwhelmingly. Were this really the case, then Bush would be struggling in the polls now in the states where there is a large Muslim population; this isn't the case though. For example, in Florida he leads all Democrats by about 15 points.