Its been years since I played with these tables, so I went out and retested this to refresh my memory, its the Aircharge mult tables for calculated load. They need to be adjusted along with the load at WOT table if you want desired load to not be limited. veeefour is right the maximum load tables don't seem to do anything. I've just told people to modify both because its been so long I forgot which was which.
Heres a stock pull vs the only change being the maximum load tables set to .65. Zero difference between the two.
Stock vs .65 max calculated load list.jpg
Desired load is governed by calculated load via IPC. This is what happens when you set aircharge multi to .65. airload and desired load follow each other as normal below the new limit. Calculated load (based on throttle airflow) raises much faster than throttle angle (based on DD table torque values). Once the ECU figures out they don't just disagree, but the airflow is much higher than the throttle angle in the TB model says it should be, you get a wrench light. Load breached .6 (.65 * the load at WOT aircharge), calculated load is at 100% and airflow was increasing. The stock system acted very quickly to a hardware problem. Desired load and all the torque based system is then set to FMEM bypassed and the pedal still uses calculated load to get its airload and engine torque values. You could just disable all the IPC and increase the allowed torque errors, and that is what most people do to stop wrench lights, well before the Whipple and Roush files were decrypted. This is just to show calculated load (the thing most people ignore) is the IPC and more trusted. The math equations modeling the physical hardware and MAP are more trusted than the air sensors or the position sensors which the desired load and TQ rely on. You should make sure these models are accurate before you just dump a bunch of error into the TQ-load tables. They are ment to take up some of the error the models inherently have, but not a whole lot like adding more than double atmospheric pressure does to the SD section and hence TB model and injectors flow model. Next to the MAF transfer, SD is the second most important model in the whole system. You can only correctly tune it with good MAF transfer, injector data, TB model, and a MAP sensor.
Line is where it light the wrench light
Aircharge mult .65 wrench.png
Operation with out the desired load/ desired torque. .65 aircharge multipler limits to .6 air load fine and still reports accurate engine brake torque. Does not drive smoothly, especially at low throttle angles, where the TB model is not accurate and the load/ torque model is better.
Operating with out desired load.png