> Do you know about Boltzmann samplers/generators, and how they compare to
> your "fueled" generators?
No. I made a mention of my deep ignorance of probability on combinatoric structures in the documentation:
Remark: I have no idea what the distribution of terms generated used this technique is, and whether it is "uniform" -- probably not. I only found it very useful in practice.
(The wording should be changed to "surely not [uniform]". If you can't prove it's uniform, it's not uniform.)
I think I don't have enough (time weighted by interest) to learn about this stuff right now. But feel free to send me emails to convince me that it's fascinating, or patches implementing better approaches.