Scientific Progress

[Torsten]

I think the issue here is that we are thinking about falsification at different "levels". I certainly agree that specific claims based on your model can be double-checked by people with the necessary skills. If your claim is that you created a system that mimics what a human could do and results in something close to Voynichese text, then both of those can be checked. And certainly the claim about the statistics could be falsified; if someone were to find significant differences, those could be pointed out.

However, if you were to claim "they did actually use a technique similar to this to make the VMS", there is no way that can be falsified or confirmed. The merit of your system is that it shows that it is humanly possible to use a form of "self-citation" to get Voynichese-like text. But I don't see how we could ever come to the conclusion that the MS was indeed created this way.

In short, your algorithm itself can be falsified. But as a Voynich solution, it can't.

It seems that you are the one moving the goalposts by introducing "absolute proof" as your definition of "falsification". First, in my view, the danger of using such a definition is that it allows you to accept hypotheses you favor as plausible and dismiss those you don't like by claiming the "ultimate absolute proof" to convince you is missing. Moreover, can we ever have absolute certainty about anything in the Voynich world? 
Second, please note that the title of our paper states, "A Possible Generating Algorithm of the Voynich Manuscript" [Timm & Schinner 2020]. In other words, we are simply arguing that it is the most plausible explanation known so far. It is possible that someday someone may present an alternative generation method that can also explain all the features of the Voynich text. However, because the Voynich text is highly structured as well as very repetitive, this method would most likely have a lot in common with the self-citation method.