Why would generality be opposed to falsifiability? Wouldn't having a model be more general lead to easier falsifiability, given that the model should apply more broadly?
In order to tell whether something is performing a computation, you try to find some way to get the object to exhibit the computation it is (allegedly) making. So -- if I understand correctly -- then a model is metaphysical, in the things you write, if applying it to a particular phenomenon requires an interpretation step which may or may not be known to be possible. How does this differ from any other model, except that you're allowing yourself to be sloppy with it?
If you just replace "metaphysic" by "model", "metaphysical assumptions" by "assumptions about our models and their applicability", "metaphysical speculation" by "speculations based on our models", I think the things you're trying to say become clearer. If a bit less fancy-sounding.
If the thing I understood is the thing you tried to say.
I could replace all my uses of the word "metaphysical" with "sloppily-general", I guess, but I'm not sure it has quite the right connotations, and "metaphysical" is already the standard terminology. "Metaphysical" is vague in a somewhat precise way that "sloppily-general" isn't. I appreciate the general need for down-to-earth language, but I also don't want to consent to the norm of encouraging people to take pains to write in such a way as to be understood by the greatest common factor of readers.
Here's the new thread for posting quotes, with the usual rules: