This is bound to happen once endgame PVEers are pouring into WvW.
Theoretically, with a 48 homeworlds setup, the maximum WvW capacity is 32k players at a given time across the board (2k players every 3 servers). With the undisputed success of GW2, the total number of players joining WvW during prime time will certainly go over 32k; hence, the queue and QQ. In reality, things would be worse, since the WvW demographics over all 48 homeworlds are NOT the same, resulting in longer queues for some homeworlds. For all the right reasons, not all WvW players would opt for migration.
What to do then?
Increasing the WvW capacity for each map has been officially talked about, so it's not an option in this discussion. Adding more homeworlds (not physical servers) with free transfers may not work, because, as stated above, some WvW players just downright don't want to transfer.
My proposal is dynamic map duplication (still a maximum of 500 players on each map as usual), along with a reduction to the points gain for each objective.
To achieve this "dynamics", there has to be a monitoring/evaluating system for queues, which produces data needed for homeworlds matchup. When a new matchup takes place, not only do homeworlds of similiar strength get matched, histories of queues from those 3 homeworlds are also taken into consideration. The more queues there were last time, the more duplications of maps there are this time.
For example, say, in the beginning of this bi-weekly matchup, due to the histories of queues from all 3 homeworlds, a duplication of the Eternal Battlegrounds is needed to accomodate the estimated queue. Then we have 5 maps to play on, so 2.5k players can join this matchup. The scoring for this matchup across 5 maps (instead of the normal 4) would also be adjust by 0.8 (4 divided by 5), which means 40 points for a castle, 20 for a keep, 8 for a tower, 4 for a camp.
In some cases, the duplication could consist of 8 maps (double maps, 2 Eternal Battlegrounds, 6 Borderlands, a maximum of 4k players) with a scoring adjustment of 0.5 (4 divided by 8), or 12 maps (triple maps for a maximum of 6k players) with a 0.33 multiplier (4 divided by 12) for the score.
Of course, the map names of those duplicates have to be sorted out. So, what's your take on this WvW queue?