I tested a climbing device I made a few years ago and I was getting hitch force at the hand vs how much the device was holding. I had the advantage of putting one cell between my bridge and the setup and one cell at my SRT rope anchor on the ceiling. With that I could see the overall support force spike with dynamics and concurrently see the amount at my bridge, the difference being how hard I was "tending?" the hitch with my hand. Then I could measure quantitatively what I felt with my arm/hand.
A problem I had was the noise filtering of the amplifiers which are usually designed to give a nice static digital readout without 60 Hz hum messing up the works. I also had two different amplifiers with different filtering. I massaged the filters to let in more noise but respond faster to catch the transients without rounding them off. I was reasonably satisfied in the end but it wasn't perfect.
The enforcer says 500 samples per second so it is probably limited to 250 Hz bandwidth or less on the signal, probably less like 100 to 200 Hz its a grey zone design decision. 100 Hz would catch any transient in a rope. You shouldn't get any huge waveform distortion falsehoods. One time many years ago I struggled to make F = m x a on a jig because of filtering issues.
If you've got two enforcers, maybe a rig like I described to Mark for measuring SRT basal and tip forces. If you set up for limited travel (so a load cell doesn't try to run through any device), put one just above the ground guy's grip, inline, couple inline alpine butterflies or whatever, then put one enforcer similarly right where you tie the log/weight. Since the rope makes a 180 at the device you just add the two readings (graphs) for the tip force and divide them to get the hand force amplification factor. The only glitch is the mass of the enforcer load cells, which if the log takes off quickly, will generate some force of their own accelerating themselves or the portion on whichever side of their sensor. You'll also measure accelerating the rope. If you do a calibration run with a block, you'll also measure accelerating the pulley and the pulley friction/rope bend friction, so use a magical bigger radius pulley thats light

I.e don't use a huge heavy steel pulley. If you use a 2" it might give you rope bend force comparable to the bend radius of your gizmo. You could do some extra figuring or ciphering with that info perhaps. Yeah, I've been watching the Beverley Hillbillies lately. Comic relief.
All clear as mud? Draw a careful picture to cipher with.
Lucky guess at x5! I can see it reasonable to vary x5 to x10 etc. Makes sense.
Edit - I just remembered that I also had a string on a pulley/potentiometer so I simultaneously recorded my descent onset/travel/stop to identify breaking initial hitch lock, then dynamic and then how the stopping action behaved. Also, each enforcer is independent or can you synchronize the data? You could tweak or snap the line to make a synchronization spike in the recorded data before the data run. Maybe whack the rope with a stick.