http://www.sfgate.com/bayarea/matier-ross/article/South-San-...
While basically matching your numbers for the new routes, the article also mentions that the effective subsidy for the older routes is about $15 per ride. I guess part of the question is how to distinguish "start up costs" from "operating costs".
It would be interesting to see if they could build a floating train somehow (trust me, this is the kind of project that gets you promoted at Google :-) basically float a monorail track between up and down the bay such that you got the energy/operating costs benefits of rail setup without the huge land/tax/easement costs of a land locked rail line.
I still found it hard to believe. So I took sfgate's numbers, and did the math.
100,000 riders a year? That's only ~270 riders per day. For an assumption of 8AM to 10PM, it isn't a hundred passengers going across the bay in per hour – it's a grand total of 19.
And for that, they built a $26M ferry terminal?
No wonder it costs so much.
BART and then Uber from 16th st or something seems like the only semi-viable non-self-drive option. Pretty expensive, though.
The idea of more bridges has been brought up but met with strong "not in my backyard" resistance. A short bridge from Bremerton to Bainbridge would make a lot of sense (and allow two ferry routes to be combined) but the wealthy Bainbridge Islanders are against being connected to the less well off Bremerton.
http://sanfranciscobayferry.com/
In Seattle it makes sense because it's between islands, some of which have no bridge access. In the Bay Area, there are bridges directly on many routes, and aside from a couple of tourist places, everywhere the ferries run is at worst two bridges or two segments of ground travel away. (In Seattle, the transit authority is actually fairly competent too, by comparison)