RealClearPolitics has a report on polling of people who have already voted in New Mexico, a sort of postal vote exit poll. Current estimate is 60% Republican 36% Democrat. The same sort of polling in the state in 2008, found that although only 10% of voters had cast their ballot early, yet the exit poll disproportionately favoured the eventual winner. RCP speculates that this might be repeated next month.
1) I'm opposed to the publication of any exit polling of any kind before the LAST polling station has closed ANYWHERE and the last postal ballot received. This is because this type of polling can be used to manipulate the result: you tell people in the state of Washington that the Republicans have lost nationally, therefore a million Republicans don't vote in Washington, giving a cheap victory to Democrats. The same would happen with Nevada if the exit polls told the opposite story.
2) I think it would be fair to consider the early voting as a sign of two influences - the "enthusiasm gap" between the two parties and the relative strength of local party organisation which may be helping supporters obtain postal ballots by sending reminders etc. Either way, unless a lot of Democrats read this and get fired up, I'm going to assume that the exit poll indicates a Republican victory in the New Mexico Governor's race.