Several things.
1) Using SMT2 vs. SMT4 in AIX is not encouraged. The dispatch algorithm with SMT4 is better
2) What ASE is telling you is that it is using 100% of the cpu cycles given. Not sure where you are measuring CPU, but one aspect could be the entitlement for the LPAR if measuring cpu from the hypervisor. If from the OS, then it is possible that some other process (and there are tons of them in an OS) is running instead of ASE and it has a bit of idle time.
3) I think you are trying to solve the wrong problem. Instead of saying "I have 32 engines buried at 100% and I have another 30% in the box to give it" - you should be asking "why do I have 32 engines pegged - can I fix that" - instead. Reason why is that usual high CPU utilization is driven by a number of factors and increasing CPU resources only makes it worse. ....and once you get that extra 30% and you still are at 100%, when then????