Peter, thanks, "lozenging" was the word I was missing.
The amount of engineering effort you put in your Maximus design is impressive, I'll read the whole thread later.
In my simulations EG filling is even more effective than the partial bracing, and it also provides the much needed vibration damping - a hollow steel tube rings like a bell.
Closing the ends is better than nothing, but much less effective than filling to prevent the lozenging.
So I believe my initial point still stands: filling the gantry tube with EG is an easy way to stabilise the cross-section with the added advantage of vibration damping, especially in a fixed gantry setting where the extra weight does not matter.