Sorry, can't provide comparison with an older MTH Mikado for flanges. Would not surprise me if that was a factor. These are pretty deep flanges, and their edges strike me as kind of sharp, so grabbing and climbing a rail may be more likely than for a different edition of the same chassis. Other than GRJ with his 2014 thread, I hope there are more case reports coming (with solutions would be ideal!) No one I know of locally who could provide a loco to compare.
It climbs the point rail on that O72 switch, but I can't feel or see any roughness or defects in that rail to explain it. If other locos do it, we'll for sure get more curious about that. The point rail has a sharper edge than the formed ordinary GarGraves-type rail, so it makes sense from that standpoint.
Center of gravity was found by balancing the loco by itself (no tender) on the 3rd drivers on a craft stick placed crosswise, which caused it to tilt decidedly onto the rear drivers. Not the least iffy, but also no more scientific than that. Likewise with new weight, same test placed it decidedly on the 2nd and 3rd driver sets with the same test process and two craft sticks. No further information to offer. Others may want to do this in a more precise way to satisfy themselves, no argument, but for me it's good enough and I'll leave it there. I'm not seeking arguments about the process, just hoping to find others who have had the problem and, if solved, how they went about it in actuality. It's clearly a rare and quirky issue, devilish to duplicate, but Lou indicates it is not unique to me. I'm personally done with it with reasonable certainty, but curious about other experience solving it in someone's actual practice. If someone added sheet lead or other material, enough to solve it, and saved the smoke feature, that would help somebody else for sure.
