That implies your max used is only about 80% which isn't brilliant but I wouldn't expect it to be as bad as your spinlock output shows. Is it at the HWM when you have the issue?
I'd suggest running that script I posted during the times when it happens and also look at the 'MemUsageKB' field in monProcessActivity to see if it gives any clues as to who may be using the proc cache.
It is possible that the large statement scenario (resolved with TF 7790) is happening, but the stack trace outputs should confirm. You should really pursue this one through tech support if it is causing you business impact.