Very good summary, Steve. I would hope everyone who reads this thread will have a better understanding of the variables involved and the need to understand the default settings. Obviously, any program that provides for user changes can cause problems when designs are shared and different values are used. From what I've seen here, I have no doubt all the various examples posted will work and the only real problem has been using different values when trying to replicate what's been posted. I guess that's why I prefer default values be set to the accepted tolerances so we're all working with the same values, at least by default. Unless someone is working on something very special with extreme tolerances, I can't see anyone changing the defaults in this area.
That said, I have no problem with the SCARM default of 2mm, but it appears that may be forcing folks to look for combinations that wouldn't be needed if the accepted tolerance of 3.8 was used instead. Some folks understand that "close" is good enough and the track will fit in reality even if it doesn't in SCARM, but others pursue a design with perfect alignment. RR-Track requires alignment in order to use the simulation feature, so anyone using RR-Track has to make sure everything fits. Of course, RR-Track uses a predetermined value, so I suspect those of us using it, do that anyway unless just roughing out a design.
Now, it might be as simple as noting the accepted tolerances for each gauge in a ReadMe file and encouraging the user to set them with that in mind. Of course, my preference would be that if I select an HO gauge track, the value would be set to 2 and if I select O gauge, it would be set to 3.8. The point is that when someone posts a design, someone else should be able to replicate it. Then too, I see all the layouts (and tutorials) being posted for SCARM and now have to wonder just how accurate they are using 2 as a default. I know they all work, but how much would those that use a lot of smaller track sections change if the value was changed from 2 to 3.8. I hope to get time today to rework my RR-Track design now that I've reset the join error value to 0. Hopefully, all the track will still fit together, I guess I'll find out soon.