I've concentrated a lot on getting my turnouts (normally the greatest source of derailments) working well.
I learned a lot about turnouts, back to back, gauge, etc. Let's just say I started learning when I could not back a 7 car train on my inner loop (which is almost all 10' curves)... after I learned a lot about flangeways, wing rails, flange depths, widths, back to back, etc. I can now back a 45 car train on the SAME track.
So, one thing I learned was about flange bearing frogs, mostly exemplified in LGB turnouts, and what happened if you ran wheels with flanges that we not as deep, clearly they now did not ride on the flange through the frog, and had to depend on back to back, and wing rail flangeway width, as well as the guard rail flangeway widths on the stock rails.
In a nutshell I came up with what works and does not look toylike, and allows me to use a set of standards. Somewhere along the way the NMRA updated it's standards and also pretty much adopted the G1MRA standards, albeit with a bit more "latitude", mostly to accomodate the more lax standards by a number of manufacturers.
The bottom line is that by following these standards, my railroad runs much better, and I can have a mix of manufacturer's products. Now, yes, I run much longer trains and demand (apparently) more reliability in running.
But the bottom line, adhering to these standards, re-gauging wheels, eliminating improper flanges on wheels, setting proper flangeway widths, makes my layout run exceptionally well. From how well the G1MRA standards work, I thought that Europe, or at least the UK would be the "Mecca" for adoption of these standards.
But from the responses here, one would be led to believe otherwise, and I am somewhat surprised.
That all (yeah I know, that's enough ha ha!)
Greg