I wonder if some of those rails are too high. Some years ago, a friend brought several of his long-wheelbase brass steamers to run and photograph on my layout, but several of them stalled on some Atlas code 83 turnouts.
I had never had an issue with any of my plastic steamers, even the ones with a reasonably long driver wheelbase.
I asked him to leave the offending locos so that I could check them out, but I couldn’t find anything wrong with them until I manually pushed one through one of the problem turnouts, and felt it bump up a bit as it entered the area of the frog.
It turned out that the frog rails were somewhat higher than the rest of the rails, and the loco, when running at a fairly low speed, would have the front drivers lifted as they encountered the high rail, while the back wheels were still in contact with the powered rail, allowing the loco to continue running onto the turnout.
However, the rear drivers eventually got lifted too, and because the springs in many brass locos are very stiff, the loco would stop, as its driver-wheelbase was pretty-much centered on the unpowered frog, the loco balanced where it couldn’t get power.
With the track power off, I put my scale rule, on-edge, across the rails and moved it over the turnout, where the too-high frog stopped it. A few passes with a not-too-coarse mill file took care of it, and I found another four or five offenders as I went around the layout, checking them all.
On re-testing those brass locomotives, none stalled or even hesitated on any of the turnouts - initially a not-readily-diagnosed problem, but one which was very easy to correct.
While I have some turnouts from Shinohara, Peco, and Micro Engineering, the majority of th