Experiments with folding to superclass, or basic configurations

[ilya-klyuchnikov]

There are variants of traditional supercompilation (including Turchin) when if current configuration is an instance (but not renaming) of some configuration, then folding is performed. Some people say that in this case we may loose an opportunity for specialization. Obviously, it is possible to make multi-result supercompiler with the only point of multi-resultness: folding + further driving when strict subclass is encountered.

Experimental results should be quite interesting.