so non-trivial evidence would lead me to put more certainty in a 'position'.

Because this particular test is getting a lot of (possibly earned) flak right now, I think it would be useful for people to actually SEE this test in action. I don't know if this is considered "trivial", but to me watching these tests is decent evidence that it is testing Theory of Mind as much as, if not more than, language skills.

Here is a TED talk on the subject. The entire talk is on the subject, but the part with the experiments is at 3:55- 7:30 . This experiment goes on to show that event though Theory of Mind is understood at age 4, that it's not until age 7 that it is used to decide moral judgements. (If the pirate ate the wrong sandwich because his got moved without his knowledge, is he being naughty/should he be punished?)

In this video the three year old is at the stage where she realizes that she USED to think that crayons were in the box, so she is starting to learn Theory of Mind. She still gets the second question wrong though. This is because this isn't a concept that one day you don't have, and all of a sudden you get it. Learning ToM is a process that takes time.

Finally, here is one that shows not understanding other's point of view.