Well, the photographic evidence would seem to imply that she does in fact have enough power to stringline the consist. Keep in mind, it's 2 foot gauge, so center of gravity easily becomes an issue. Once the start tilting, not much force needed to keep them going.

Anyone who's spent much time around small scale live steam, like 7-1/2" gauge, has seen this type of thing happen on a routine basis.

Came here to say this. Spaced far to wide to contact the flanges while they're still on the rail. The goal of this one is just like a guard rail on a bridge. If you drop off the rail, you bounce along between the two rails, guiding you in a general direction.

Why only part of the curve? Could be sharper there, or could be that the terrain is steeper, meaning a derailment at that point would be more likely to fall down a hillside.

I noticed in the video that there were no center pins between the truck and the carbody. If they had been using locking center pins, the weight of the trucks would have figured into the CG and the cars might have stayed upright.

It was suggested to me that over the years the trucks on the cars were never turned to run in the opposite direction and uneven flange wear could have been a contributing factor. The term "pizza wheel", for how much flange there was on one side was used.

It's plain to see in the pictures... "thin flange" is not an issue here.

We'll probably be surprised to learn which car actually derailed first...

I'll only be surprised if it isn't the derailed car that's furthest from the loco. Classic stringlining it would appear.

.

I've watched a lot of video of the Romney, Hythe and Dymchurch over the last few years, and one of the most striking things about it is that they are running 1/3rd scale locomotives on quarter scale track gauge, with cars that are a little larger than that, at speeds up to 30 mph, on stick rail that in some places, under telephoto view looks like the bad old days of Penn Central, and they don't have problems like this. I don't think the problem was necessarily over-size equipment for the gauge. So, probably something happened that we will have to learn about in an investigation. Hopefully it will be independent, thorough and professional.

I’ve been a theme park enthusiast myself for a while and I can’t say I’ve ever seen this happen

I did catch an interview, which now I can’t seem to find buried amongst all the stories, of a witness that said a car was bouncing up and down right before it went over. That says derailment then tip to me. She may have been in the last car and was referring to the one in front of it - don’t recall The last car didn’t tip. If someone runs across that interview, post the link.

A couple of issues that I wonder about: does the length of the train or the length of the individual cars have an effect on stringlineing? And the effects of scale on speed, is a 2 foot gauge train going 10 mph the same as a standard gauge train going 20 mph? Side bearing clearances scaled out to standard gauge could be excessive especially when combined with larger than scale width of the cars. Isn't ASTM trying to come up with standards for amusement park rides?