Hmm, I see what you mean now.
I think "closing price" here would be not really fair, as any particular time you'll pick up for this calculation will be inevitably favoring some markets that fit their daytime trading activity nicely and neglect others.
A volume-weighted average over last 24h seems much better to me here. Because it will include activity on all global markets, which ranges and differs a lot during the day, and won't favor particular markets above the others, but will consider trading volume only, which is measurable and unbiased parameter. And it can be shot at any time during the day, particular time is not important in this case. Midnight by GMT would be handy reference point probably.
Closing price is what Coindesk does, but I'm not sure it makes a lot of sense for a market that doesn't close. It you had to pick a single time for a daily pricing 'fix' then noon would make more sense IMHO. Personally I think a 24hr average, midnight-to-midnight, would make a lot of sense (and midnight GMT is as good a choice of midnight as any, although I don't care that much).
I'm not really wedded to details here; I'd just like to see more people publishing tables that say Bitcoin was worth this much on this particular date. (If multiple such sources emerge I will probably just use a simple average of the published indices unless/until one emerges as the de facto standard.)
roy