I do believe, however, that a C++ Fluidsynth may not be quite as easy to absorb and apply as a C Fluidsynth.
Once the C++ API is mostly stable I would like to provide C bindings to it. They should be fairly easy to generate from the C++ headers, although to my surprise I haven't found a C binding generator for C++ yet (anyone knows one of these?).
Anyway, I believe these days almost everybody is familiar with some language with OOP features (Python, Java, etc), so learning enough C++ to use Fluidsynth shouldn't be an important barrier.
2) I hope that if the "Qt way" is embraced by Fluidsynth that there will
always be a great amount of freely available documentation and other
information about things Qt. In looking at Trolltech's web presence
I'm a bit suspicious of their desire to fully and freely support the
actions of independent and indigent programmers like myself.
Indeed Qt's API is widely recognized as high-quality and my programming style and some of my ideas have been partly inspired by it. However, the only part of Qt I'm using in this project is QMake for the build system, which would be very easily replaced should it become unavailable (which I highly doubt anyway).