I cannot see how this is relevant at all to TeXmacs.
On the other hand, you could think about
a more conservative approach, namely,
translate the desired scripting languages
to Scheme.
This is a big deal. IIRC, that was one of the stated initial goal of Scheme. And I am not aware that anything happened in this direction.
Especially: Python and _javascript_ should
be considered. Python because the combination
C/C++ plus Python is the approach of choice
when people need to perform resource intensive
scientific calculations (e.g. simulations).
It would be more sensible to port TeXmacs to another scripting language. This would avoid a dramatic increase in the overall complexity of the system.
This being said, I don't know when the
resources will be available for tackling
such tasks...
In my completely irrelevant opinion, it would be a much much better idea to fix the user interface.
If any work should be done on the scripting aspect, it should rather focus on adding an automated test suite and either documenting or removing metalinguistic complexity such as the DRD system.