We're in the process of implementing alternate solutions to help assist the lower population realms. The problem here is that Agamaggan is not really what would be considered a "low population" realm, it's not even in the bottom 25 realms. While it's not at capacity (it's at around a 55% nightly pop) there are quite a few realms below it that we would need to focus our attention on first, such as Coilfang (around a 30% nightly pop) and 10 or so other realms like it.What that tells us is that the most problematic servers are only one-third full during prime time, and many medium population servers are only half full, which Blizzard doesn't even consider as problematic. We know that Blizzard is aware and working on a solution, but they aren't telling us yet what the solution looks like. Makes me wonder whether they will come up with anything better than the classic server merger solution.
Just to compare there are technically around 200 realms that aren't actually hitting capacity (no queues at any time for a month+) out of about 220 total US realms. While we would like all realms to be near capacity, there is a middle ground that still allows an extremely functional and fun play experience. A lot of the realms live in this "middle area" of capacity, and they have extremely lively economies and interactions. Before the population cap increase with Burning Crusade they might even be considered "high population".
One Blizzard specific problem is that World of Warcraft has much longer periods between expansions, and it is right after an expansion comes out that many people resubscribe. So they can't go overboard with merging servers just to see them overcrowded once the Wrath of the Lich King comes out next year.