Triplex, that was my thought as well. The thread I linked (if you read through it, and assuming it is scholarly work) stated that at one point on the Pennsy tests, they had the J up to 115... which is when the piston valve seized in the valve cage. 110 for drivers that size is very impressive, 115 is astounding. Maybe 120 wasn't impossible, but I'd think the machine would be taking an astounding beating at that speed. At 110, the drivers would be turning 528 rpm (assuming no wear on the tires) - 8.8 times per second with reciprocating gear going from dead stop, reversing, to dead stop again at twice that rate. I'd like to figure out what the piston speed is mid stroke, but I don't have the math for that. Wow. It's amazing they didn't throw parts everywhere.
I don't have the citation, but I've read that N&W took a different tack on their counterbalancing scheme to allow high speeds without sacrificing acceleration / hill climbing advantage that small (for passenger engines) drivers allowed. One of a counterweight's tasks is to offset piston thrusts to keep thrust-induced yaw under control. The downside to correct thrust counter weighting is a wheel that is more out of balance vertically, leading to severe rail hammering as speeds increase. N&W chose to increase stiffness of the lead trucks lateral motion to control yaw, allowing the overall counterweighting to be reduced. I have have no idea whether this decision was to allow higher speeds or to reduce track hammering, but I'd suspect the former. I imagine that the extra stiffness had it's own disadvantages, but those must have been manageable.
Fluesheet