I think I understand what you are trying to accomplish. Using IAT to predict and adjust fueling and timing may prove to be difficult due to the inherent latency in thermistors that you already have discovered.
Folks have been adjusting timing and fueling on boosted engines for a long time, so they must be using a a different logic and type of data to do so.
FP regulators with pressure compensation have worked well to adjust fueling during boost transients. Detonation detection requires a bit more sophistication.

Pressure-Temperature relationships are fully developed and can be used. It may take some testing or convincing that pressure data and math is better to keep ahead of IAT transients that may need timing and fuel adjustments to stave off detonation. It would also take some testing to determine whether the transient is necessary to adjust for or the new steady state that one needs to address.
It still boils down to the communication rate the ECU has with sensors, which I do not know for the Holley ECU. Communication rates are not to be confused with the internal processor speed as that is for calculations and performing instructions. I know that general UCM (Universal Control Module) query I/O rates are around 100Hz and sometimes up to 200Hz if you find the right one. (CAN is usually pretty slow unless it has optional high speed capability.) This information is usually found on the UCM print from the vendor or in an SRD (System Requirements Document) and is not shared openly. I'm not an EE so I'm at the end of my knowledge on these types of things at this point. Anyway, good luck in your endeavors and hope you avoid making aluminum ash trays.