This is pretty much how I feel.
I think Torsten's papers are full of very good observations that should be read by every person who is serious about studying the VMS.
You can count on two fingers the number of people who are trying to explain not just the composition of a token but how to get from one token to the next. But... I can think of other methods that might result in self-similar text that do not rely on autocopying, so I don't necessarily agree with an autocopying conclusion until those other possibilities are eliminated.
I think Torsten's papers are full of very good observations that should be read by every person who is serious about studying the VMS.
You can count on two fingers the number of people who are trying to explain not just the composition of a token but how to get from one token to the next. But... I can think of other methods that might result in self-similar text that do not rely on autocopying, so I don't necessarily agree with an autocopying conclusion until those other possibilities are eliminated.