To answer the initial question, there would need to be a hardware upgrade to convert to stored value cards (SV): the current swipe readers (*) are read-only, while you need a read/write to implement SV. If you're going to that length, at this point I wonder if it wouldn't be better to step up to a contactless card instead.(**)
The fare collection study had two parts: one on policy/fare structure and the other on technology. The policy side (I served on an advisory panel to it) was rather interesting because it shed light on management's perspective: while the passengers wanted better integration of fares (e.g. transit/railroad) and a lower transfer charge, management wanted to reduce the number of different fare instruments SEPTA sells, particularly for the railroad. That's one reason they floated a reduction in the number of RRD fare zones a few years ago. The bottom line was that adopting a new fare collection system was attractive to a lot of people, but it wouldn't have the impact it did in New York(***) and it would upset some people who find the present system advantageous (like those who make 10 one-seat trips per week and need only pay for a bag of tokens instead of a pass). But mainly SEPTA didn't have the capital money to pay for a new system, and they didn't want to look around in the budget to try and find some money.
There is indeed a counting room, and last I heard, SEPTA contracted out token packaging (don't know if they still do, but at one time, it was done by a workshop for mentally-disabled people, which I think is a win-win). Don't know what the current count is, but around the time of the aborted fare increase and resulting token shortage, SEPTA had about six million of them. At 5 g apiece (about what a nickel weighs), that would be 30,000 kg or 30 metric tons.
And yes, Jeff, no fare technology is going to be effective if you don't make it easy for people to buy the fares.
*--remember that at the time SEPTA installed them (late 80s), SEPTA was actually out front on fare policy in several ways: widespread pass use, the high cash fare and steep token/pass discount, and the swipe-reader turnstiles. Since then though, technology has improved markedly, and SEPTA is now behind the times(****).
**--then again I'd also give a lot of consideration to a proof-of-payment system: you could eliminate transfer charges entirely this way as well as speed up boarding and alighting, plus other benefits.
**--Everyone wanted to replicate the Metrocard success (huge increase in ridership), but many forgot why Metrocard was a success: it took the TA in one great leap from a 1920s fare structure to a 1990s one. There were no unlimited-use passes, no opportunities for discounting, and transfer passengers had to pay two full fares. The benefit wasn't from technology--it was from modernizing the structure of the fares.
***--at least they haven't adopted some fare fads like the flat fare with no transfer discount (i.e. pay two full fares for a two-vehicle trip). The accountants love this system, and it was recommended in the Phoenix report, but it's economically inefficient because you're increasing the fare penalty for transfers to 100% on top of the time penalty. For some systems with radial networks (e.g. Baltimore), it's feasible, but not for a grid system like SEPTA's where the most efficient way to move people is with buses feeding rail.