For anybody interested, here's an update from today's comp.lang.c traffic commenting on the testing performed against Ben Pfaff's PRNG implementation...

----------------------------------------------------------------

It seems to me that Ben has done a nice job here.  His PRNG passes
George Marsaglia's "diehard" test with flying colors.

http://stat.fsu.edu/pub/diehard/

(Last time I checked, my local standard rand() did _not_ pass; the
Mersenne Twister does fine).

In case you're interested, Ben, I've put a little archive of the
diehard test suite as it applies to your code at

http://www.ecn.wfu.edu/~cottrell/pfaff_prng.tgz (60Kb)

-- 
Allin Cottrell
Department of Economics
Wake Forest University, NC