THC not activating

uname -v: #1 SMP PREEMPT_RT Debian 5.10.84-1 (2021-12-08)
My install is a probably a one-off. I am fairly sure I did everything the hard way. I am running Debian 11 with Linuxcnc 2.9. You and several others helped me many months ago get all the right things in the right places to make this work.

Maybe I just don't have my head wrapped around the terminology.
My assumption is that MAX_VELOCITY would be the fastest feed rate possible.
MAX_ACCELERATION would be how fast it gets to max velocity.
However, when leaving MAX_VELOCITY alone and turning MAX_accel up to 100 IPS2 in the ini file, the movements are super fast with a 200IPM setting in material tab. Change the material tab feed rate to 50 and the speed is exactly the same according to Halscope. Is my gcode lacking? And why is there a button to disable the Velocity Anti Dive if it doesn't disable it? Disabled anti dive would be super helpful in troubleshooting. I feel as though I have gone from one extreme to another with velocity and accel settings with no real indications of where to go. I don't remember having these type issues in the beginning other than learning how PIDs work. Definitely not saying that I know what Im doing, just that something, somewhere has changed and I cannot seem to find it. On a positive note, the last OS update did fix my issue of the wifi turning on and off randomly!

If you have lots of short segments and corners, physics means you will never get to max velocity regardless of  your cut speed.
I don't think the MDI uses max velocity and speeds either.
 

-check the machine velocity and cutting feed rate
-cutting feed rate can not be higher than max machine velocity
--
From what i gathered, that is the only issue.

It is most definitely fixed this time! I checked the velocity by cutting 2 inch lines with x-vel-cmd in halscope running. The velocities were spot on with commanded vs. actual.
The acceleration however was way off. I adjusted the MAX_ACCEL in the ini file match commanded vs. actual and I am actually cutting better than before! I also ran arc voltage that helped me see that 1 out of 5 pierces, the THC delay was a bit too early causing erroneous sampling. I am assuming that the acceleration problems were due to gantry weight, friction, driver and stepper limitations, etc., something Rodw mentioned. And thank you Tommy for the idea of checking! I know verification should be the first step in troubleshooting, but I seem to run around with blinders on sometimes!

A bit of testing always helps a lot!