But there are many many more players at the lower limits that more than make up for that.
Erm... no.
Let's use STTs as an example. A $1 + 20c STT generates $1.80 revenue for the site. An $10 + $1 STT generates $9 in revenue, and a $100 +$10 STT generates $90.
The thing is, none of those STTs is substantially different to any other in terms of running costs. Each one needs the same amount of server processing power, the same internet bandwidth, needs the same support staff and a bunch of other stuff besides.
Let's say those things add up to 90c per STT*. I don't work for a
poker site, I've got no idea what the actual figure is and 90c is probably high. But it'll do for the sake of argument. If you take that as the figure and reduce rake on the $1 STT to 10c, all of a sudden you're only breaking even - why bother offering the tournament?
If the site cut the rake to 10c they'd need to double their volume in those games to make up the lost revenue - anyone really think that's going to happen?
Lastly, you're ignoring probably the most imporant fact: The high volume of players you're referring to proves that 20c rake on a $1 STT is a price the market is willing to pay. The market has spoken, what's the incentive for the site to change?
* for anyone who's thought that hard about it yes, I'm aware that any figure would come down slightly at high volumes as fixed costs are accounted for and only the variable costs remain. Again, I'm just illustrating and the point remains that lower stakes games contribute much less to overall profit despite their similar costs**
** OMFG I can't believe I just made a semi-serious business post on here