The interesting thing for me is that, unlike many of your fossorial reptiles that prefer sandy soil, I normally find these in harder, but moist soil. This could be because the tunnels the ants use are probably stronger in that kind of soil, whereas the tunnels would collapse in sandy soil. I suspect these snakes make more use of existing tunnels than actually excavating tunnels themselves, although I don't doubt that they are capable of this. Filling up and blocking a complete tunnel might also be a great defensive method as that means only the head and tail is exposed to the worker ants. The head has a big strong rostral scale in front and the tail often terminates in a spike, so that sounds like some good defence from my point of view! But this is my own theory... so please don't quote me on this!
The general areas that I know of where M.shlegelii were found are all moist savanha, and the 2 specific spots I know of are not sandy at all. So in all likelihood M.schlegelii uses the same soil type as R.lalandei and A.bibronii...