I looked at the mph number for the S/X on the 240 volt connector as the starting point:
The S/X 240V connector doesn't even come into play. Tesla explicitly says in the press kit how fast the Model 3 charges on 240V/32A. There's no ambiguity.
If you want the answer to why you're coming in with low numbers, it's because that's an adapter for a NEMA 14-30 but you can't actually draw 30 amps from a 14-30 continuously, the max is 24A because of the 80% rule. So the M3 Press Kit's figure of 30mi @ 240V/32A represents 32/24=133% as much power relative to the NEMA 14-30 adapter. The NEMA 14-50 can likewise provide no more than 40A continuous, meaning the M3 example is for 80% as much power. Versus the 5-15 adapter, they cite 12A, so the M3 press kit figure is 32/12=266% the current, and at double the voltage, 533% the power.
If we scale the figures on that page by these percents, we find that the "S/X" (which don't actually get the same range per hour, but we'll ignore that) get a nominal 133%*16=21,3mi, 80%*25=20mi, and 16mi, respectively. The press kit's citing M3 as getting 30 miles at 240V/32A corresponds well to the efficiency difference between M3 and MS / MX.
In the latter case, it comes in a bit under the efficiency difference, and I think I can guess why - at such low powers, the fraction of the energy taken up by parasitic draw becomes more significant. But since they're charging at 12A, when you can actually charge at 16A from kitchen/bathroom/garage sockets, the fraction of energy taken up by parasitic draw will be reduced. On the other hand, I was only crediting 15A draw, when technically you're perfectly fine to draw 16A.
Maybe assuming 5.5 - 6mph would be better due to parasitic losses at low charge powers.