Higher RPM combines well with smaller diameter platters (to keep mechanical stresses in check) and, to some extent, lower areal density (wider tracks) to make for incredible access times. The tradeoff is low capacity. This works at Enterprise level because capacity is meaningless if your disks can't keep up with the workload.
Desktop access patterns are localized, making random access less of a priority. A single user would have to work at it to overload a single 7200 RPM drive, so its performance will be technically sufficient. A single user can use more than 300GB, though (largest VelociRaptor), so capacity is more likely to determine suitability.
Thus high density, large-platter 7200rpm drives. Two 1TB Caviar Blacks can be had for the price of one 300GB velociraptor. 6.67x the capacity per dollar for only slightly less desktop performance. Competition in the 10k arena would improve things somewhat, but the move to 2.5in to get 10k cheaply cripples capacity in a way that the added performance can't overcome.
7200rpm seems to be the sweet spot between desktop performance, capacity, noise, and cost to manufacture.
but, it was odd. Western Dig. had this SATA product line that was very popular and very profitable, and no-one imitated it.
To enter, you're looking at splitting the very small enthusiast market. It's not even guaranteed to be in half, as a third or fourth manufacturer could see fit to enter. (Seagate, Maxtor, Hitachi, Fujitsu, and HP all have 10k drives.) And you won't see the profit margins that WD is seeing in an uncompetitive market since you'll be the entry of competition.
It's probably not worth the start-up costs.