The really interesting number to me is that MTA in New York City represents 1/3rd of the total transit ridership is the United States.
That other 2/3rds includes other New York based systems (Jersy Path, Long Island RailRoad, Staten Island Railway, Metro-North Railroad, etc.)
New Yorks population is ~6% of the US.
To be more precise, 30% of total ridership. So together with the other systems in the New York-Newark-Bridgeport CSA (7% of the national pop, 8% of the urban pop) - MNRR, LIRR, SIR, PATH and NJT - it comes to just around 1/3. But if you crunch the numbers it isn't all that surprising, given that the practicality of good transit service varies almost geometrically with pop density. The pop density of the urban areas in the NNB CSA (NYC, Newark, Hudson County, Long Island) range from 5-20K/km^2, by far the highest in the US and Canada. In the US it is comparable really only by the cities in the Boston area (4-7K/km^2) and the cities of SF and Chicago themselves (7K/km^2 and 5K/km^2), which incidentally have monthly riderships of 35M, 25M and 50M respectively. Using Boston as an example, the core area served by the MBTA has a pop 10-15% that of the "transit area" of NNB and 1/2-1/3 its density, but it still has more than 1/10 of NNB's ridership, still very respectable and the ratio is comparable. Chicago, with 1/3 the pop but 1/4 the density, has only 1/6 of the ridership. Los Angeles County has more than half of NNB's population, but with density only about 1/6, its ridership ends up being less than 15%. With much of the rest of the population living in suburbs around Californian or Midwestern cities with minimal pop density, it's not unreasonable to expect that transit ridership is concentrated in the dense urban areas.
A problem with counting the NYC subway separately from the commuter/suburban railways, though, is that it significantly double-counts, because the points where those services terminate in NYC - Grand Central (to a lesser extent), Staten Island Ferry Terminal, and especially Penn Station - are relatively far from employment hubs and most passengers must transfer to the subway, so most of the trips on the outer system would also have been counted in the subway ridership already.