But I have not the time right now to check. I am busy trying to prove
that 20 is minimal for superfliptwist. I have already found that there
is no 19 turn solution with 16 turns in phase 1. That took about 48
hours (distributed over 6 machines). Now I am doing the same for 17
turns in phase 1; which wil obviously take much longer. (And yes, I
took the precaution of allowing as the first turn only F, F2, R, R2,
U, U2 in phase 1. When I go to 19 turns in phase 1, I can skip 18,
I need only F, F2, R and R2, I think.)

in fact, you can eliminate the possibility of starting with F2, R2 or U2,
since these each commute with superfliptwist, and may be done in stage 2.
in other words, if F2 sequence = superfliptwist, then also
sequence F2 = superfliptwist.

also, you need not consider 19 turns in stage 1. by symmetry, you may
suppose the last face turned is U, which is done in stage 2.

if you use the fact that U and D commute, L and R commute and F and B
commute, then the number of sequences of length n in stage 1 grows
exponentially, with ratio approximately 13.35. if the runtime is
proportional to the number of sequences tested in stage 1, (which
may or may not be the case) that would mean testing 18 turns deep
would take approximately 178.18 times as long. (eliminating the
possibility of starting with F2, R2 or U2 would cut that in half.)

here's something you may have already considered. if your prune tables
in stage 1 consider only pairs (flip, twist), (flip, location) and
(twist, location), some search paths may be pruned 8 turns early.
(each of these pairs had positions 9 twists from start.) at the
expense of a lot more memory, you can do some pruning 11 turns early,
by storing tables for triples (flip, twist, location). you'd probably
have to store these tables in very compressed form, and divide out by
symmetries of the cube that preserve the U-D axis. it may turn out
that the overhead of processing this compressed information does not
adequately compensate for the improved foresight, but it's worth
considering.

it would be excellent if you could show that 20 face turns is minimal
for superfliptwist! even finding a shorter solution would be great!

mike