We should keep in mind though, that it's the number of contestants (?right vocab?) which actually created the "difficulties". I guess we have a share of regular voters that look at the challenge, prolly don't participate, but still vote. The more contestants we get, the more spread out the votes will be, leading to - let me put it with "less satisfying results".
Too bad that the timeframe is already short (I mean, it's good, for it allows a regular challenge), otherwise I'd suggest two rounds of voting (or more, if really necessary) where you pick your favourite each time. It gives all those, whose former favourites FOOFed (Fell Out Of Favour), not quite the feeling that they had been totally wasting their vote. And other could also change their mind, I heard that happens.
Downpoints: Takes too long for the challenge to remain a regular monthly challenge. I'd rather have something like a super tuesday, for the mapper's sakes. Also, Some might be prevented from participating in each vote, which doesn't sound fair.
Easiest solution: We need more - and I'd like spell that with a capital M - regular challenge voters, because in that case, the empirical statistics will... ehm... well, they work better the more people are around. :-)