Lousy comparison to be going back to 1911. No Boylston St. subway extension and no Lechmere Viaduct extension back then. And no signal system until those extensions went online. The entire subway--Public Gardens portal to Haymarket portal, Pleasant St. portal to Haymarket portal, and only 5 subway stations--was just a collector and looping mechanism for surface lines. Line-of-sight operation, uniform 5 MPH speed limit, and packing of those 38 ft. trains as close as they could get knowing that the speed limit was uniform.
It stopped being like that very quickly. Within 3 years the extensions opened, the signal system went online, and 2-car trains of center-entrance crowd-swallowers started operating. A more appropriate comparison would be sometime post-1919 when the subway was complete and platform extensions + a supplemental order of center-entrance cars allowed triplets to run for the first time. The comparison with that year probably still is far from accurate, but at least you're talking dispatched traffic control and trains of substantial length and heft starting to resemble a high-capacity light rail operation.
Also...this "OhRatFarts" Reddit guy can't possibly make realistic comparisons to people being moved without knowing the seating capacity of the 1911 trolley fleet. In 1911 BERy was using Type 1's, Type 2's, Type 3's, Type 4/4A/4A1/4A2/4A3/4A4's, various open- and closed-car 1890's streetcar stock, and "foreign" power from other streetcar companies who were still running into the subway.
Green can definitely do better, and we'll see just how much so when the GLT data analysis comes out. But 108 years ago is a bunk comparison. And, really, the
needs for moving riders around have evolved so much that looking anywhere many decades into the past isn't going to be a useful metric matching closely with modern needs.
EDIT: Until GLT results are published, EGE's
master's thesis on GL capacity is the closest thing there is to a definitive answer on what the modern Central Subway can handle.