Over on the DVARP (yay!) Facebook (boo! hiss!) page, a discussion is brewing over the internal layout of the M-4 Market-Frankford Line trains. The current setup involves nearly all seats facing either forwards or backwards, with a narrow aisle running through the seating areas. Posters in the DVARP thread are advocating converting to longitudinal seating (backs against the sides of the car, facing the center aisle), for the purpose of a more pleasant experience for standees, at the cost of a few seats per car. Those arguing for the status quo argue that the loss of seating capacity, especially in the off-peak hours, is not worth the ability to cram in more standees in the peak.
I have to come down on the side of longitudinal seating. Every passenger on the El starts and ends their ride as a standee. The current setup does not allow for more than a handful of standees per car, before awkward contortionism is required as people try to navigate to the doors for their stop. While I sometimes ride the El desiring a seat above all else, there are other times when I would actually rather stand than take an available seat, and my observations of other riders on the El, the Subway, and in other cities, suggest I am not alone. With the El definitely at capacity at rush hour and beyond, it’s time to treat it like what it is: the trunk line that carries more riders than the entire Regional Rail division put together, along the city’s main commercial axis. Station dwell is a function of how efficiently trains can unload and load. Speed the ride. Turn the seats sideways.
Sic Transit Philadelphia 
It’s a subway — the idea is that stops are frequent-in-occurrence and short-in-duration (ideally the opposite of a train, though commuter trains fall somewhere in between), and thus people are constantly, and simultaneously, in a state of enter-findopenseat or vacateseat-exit, and those seats are in random locations. It’s in everybody’s best interest to have those paths less congested, and it’s in everybody’s best interest to allow entering passengers to find and get to an open seat as easily as possible to minimize the number of riders standing in those congested paths because they can’t get to that-seat-way-over-there.
I suppose the problem is that Philadelphia just doesn’t have many subways and many riders take the MFL from one of its points of origin or to one of its termini, leading to their misidentifying it as a train? (Do more MFL trips originate or terminate at an end-point than do trips on, say, the average NYC subway line? I assume yes, by a country mile, but I’ve never seen statistics.)
I think it’s not so much by endpoint, but rather a few stops in the middle that most trips involve. Specifically 15th, probably 13th and 30th to a lesser extent. But it’s also no accident that there are “transposition centers” at each end of the line – there are lots of feeder routes in the Northeast and Delco that funnel riders onto the el to get to Center City.
I believe there is quite a bit of hardware and electronics located under the seats of the M-4s, such that relocating them would be non-trivial. That said, it’s the right thing to do – squeezing in and out of the seats is no fun. Personally, I’m not a fan of sitting sideways on moving vehicles (my preference is rear-facing), but I recognize that it’s a superior in-car arrangement. Hopefully SEPTA realizes this when developing the k-car replacements – they are very clog-prone with standees.
I was a daily El rider for many years until the last time I moved. Longitudinal seating makes sense for a place like New York where crush or near crush capacity can be common on any line, anywhere, anytime seemingly. Outside of rush hour, that isn’t the case for the El and during heightened periods of activity such as weekend afternoons it still doesn’t seem so big. During the MFSE Reconstruction I was glad to see that platform entry points were staggered at each station. Since a large number of people didn’t wander far from where they came in before the reconstruction packed center cars an comparatively sparse end cars were typical, especially in the morning. Now things are a little more evenly distributed. Headways are decent enough as well. Now such a set-up might make sense in the design stage of the next generation of Broad Street Subway cars which I’d guess would be sometime in the next decade if the B-IV’s stick around as long as the M3’s did.