It's not as bad as you think. It's measuring system length, not track-mile, so the Transbay Tube only comes out to 6 miles including approaches.
Also, it's a chokepoint: there's not a lot of alternatives to crossing the bay, so it should also be driving a much higher numerator as well. And it's not like other systems don't have geographic barriers--consider the crossings of the East River or Boston Harbor.
For the length of BART, taking out the Transbay Tube doesn't reduce track mileage by all that much, and consider that BART is ⅔ the ridership of the Washington Metro by mile. That's not a small gap. Maybe you could say it's slightly better than MARTA, but it's far from Philly or Chicago's performance, let alone NYC or international statistics.
BART is currently running at what could be considered its maximum theoretical capacity given conditions. What I mean is that the built environment provides only unidirectional flow of passengers: toward SF in the morning, way from SF in the evening.
With the addition of tracks to San Jose, this could change. We could start getting more flows of people where neither end is SF. That should boost ridership but of course they are adding a lot of miles.
Also, it's a chokepoint: there's not a lot of alternatives to crossing the bay, so it should also be driving a much higher numerator as well. And it's not like other systems don't have geographic barriers--consider the crossings of the East River or Boston Harbor.
For the length of BART, taking out the Transbay Tube doesn't reduce track mileage by all that much, and consider that BART is ⅔ the ridership of the Washington Metro by mile. That's not a small gap. Maybe you could say it's slightly better than MARTA, but it's far from Philly or Chicago's performance, let alone NYC or international statistics.