I think the analogy is closer to this:
http://bbs.hitechcreations.com/smf/index.php/topic,296813.msg3789935.html#msg3789935
edit: I still back multiple arenas, with fixed map-dependent caps. let us chose where to fly.
I fail to perceive a difference in the analogy. In either scenario the hypothetical business has had to make a decision that best supports it's ability to provide the best experience at the busiest of times by implementing a solution that at other times is not as popular. And in our hypothetical restaurants, we both know what the answer would be - a number of people who might desire to eat there at the off-peak times if they could get the same experience as is available at the peak time would choose to eat elsewhere. I'm sure that if (when) that happens here HTC is no more happy than the hypothetical restaurant owner. But both have a solution that suits the largest number of people.
I don't honestly understand how the fixed map-dependent caps would be better. In fact, it might often be significantly worse - if either arena has a cap of half of the daily average, then one arena might go all night on a slow night with 50 players in it total. On another night there might be NO slots available in either one.
As it is now, both reach "critical mass" in the least amount of time possible, and there's always room for another player.
I'm not arguing in favor of them by the way. I'm arguing against the 9000th whine I've read about the same issue.
<S>