Marty (last name?) had his O-scale 4-12-2s on display at Cocoa beach. He was especially proud (rightfully so) of his model with the triple Walshearts valve gear. I noticed the timing issue as well, and asked him about it. He pulled out drawings. The outboard cylinders were, indeed, 90 degrees out of phase with each other, as a conventional locomotive would be. The inner and engineer side valve motion were off only by the angle that the middle cylinder was tilted up so that the connecting rod to the #2 driver could clear the #1 driver.
I believe the Gresley valve gear as used on most 3 cylinder locomotives requires a nominal 120 degree offset. I say nomiinal, because with the center cylinder tipped up for the connecting rod to clear the #1 cylinder, there has to be a similar phase differential. This accounted for the SP's 4-10-2s to be called "Stuttering Dec's." See Robert Chruch's recent book, "Southern Pacific Ten-Coupled Locomotives" published by Signature Press. If it weren't for the tipped-up center cylinder, the pistons could all be perfectly 120 degrees out of phase, and the locomotive would have an even exhaust note (just like, say, vintage 1970s 3 cylinder motorcycles, like the Triumph Trident, Yamaha XS 750, and various Kawasaki and Suzuki 2-stroke triples).
I was very impressed by Marty's work. The two Walshearts valve gears working side by side on the engineer's side was a mechanical tour de force! It was difficult to pull off in O-scale, and discussing it with him, if it could be done in HO, it would require a master watchmaker. to do it as well.
I'll leave it to those who know more about these UP locomotives, but I'm also puzzled by the inner and engineers side pistons being so close in phase with each other, and the dynamic augment that must have resulted. They must have pounded the heck out of the trackage!
Arved_Grass@yahoo.com or Arved@I-Do-Photography.com
Fleming Island, Florida