As a guy who knows some hominids who know a rescue dog that is a brake engineer, I'm confused by most of this thread.
Some thoughts from my rescue dog friend:
0. We have 125 HP engines and ~450-800 HP of brakes. Clearly a lot of passing happens under braking.
1. We have OEM proportioning valves. Curves are published (Flyin' Miata, etc.), but none of us really know where we are on those curves, except to say we can safely assume we are often, except for initial transients, and "brushing" the brakes, and light trailbraking, past the "knee" of the curve. My rescue dog friend instrumented and proved this, once upon a time. According to my rescue dog friend, if you are going to modify your proportioning valve for legal use in a class other than SM, you should probably change the spring rather than alter the spring preload with shims.
2. The grabbier the pads, the LESS likely you are to be past or well past the knee of the prop valve curve. The proportioning valve doesn't know your decel rate, it only knows front line pressure. Depending on how grabby your brakes are, you may or may not be traversing the "knee" of the prop valve curve depending on whether you are brushing the brakes into a Kink, trailbraking into a sweeper, or divebombing into your best friend's driver's door in hopes of winning 13th place on the practice day of the Smitty McSmitterson Memorial Double Regional.
3. The driver is a "Flatspotting Prevention Controller" when it comes to braking (and also an "Anti-Loss-Of-Control Controller" in the event of rear brakes locking early and often). Thus, the earliest locking brake is what trains the driver with respect to acceptable pedal pressure, and no road racer prefers a car whose rear brakes lock early and often. Therefore, your braking performance shown in data and your GoPro vids while playing "divebomb chicken" is most likely determined by your most *overbraked* wheel, not your most *underbraked* wheel.
4. Camber and suspension matters. When the nose of the car slams onto the bumpstops and has -5 (-7? -45? what are we up to now?) degrees of camber, it affects the braking potential. Same for the rear. Induced rear toe is a thing, too. You can play around on the alignment rack to determine what happens "statically" ("kinematically", for the twitchy engineers) when the nose drops and the ass raises, but the compliance in cornering matters even more, in my opinion. This is ancient history, but the stock bushings are/were softer ahead of the rear knuckle than behind, but the "Comp" bushings were the same stiffness ahead and behind. Presumably the goal was to reduce the induced toe-in under cornering (and the induced toe-out under braking).
5. Pad "grabbiness" (coefficient of friction) is a function of temps. Are your pad temps the same at Gingerman as they are entering T5 at Road America? Of course not.
My rescue dog friend's opinion: 71.3% of brake problems can be fixed by simply giving the car what it is clearly asking for. If YOUR car locks the rear early, and it is holding you back on any given weekend and track, just fix it. If YOUR car locks one corner of your 52.3% crossweight car as you enter the sidehill braking zone of the Mid Ohio backstraight, and it is holding you back, just fix it. Different pad compounds is an easy and reversible way to do this. As I've told many who go to Grattan, if your RR wheel locks early into the backstraight hairpin passing zone, FIX IT. Put on a lesser pad, a stock pad, flip an outboard pad metal to metal. If the laptimer is happy, it is the right decision.
Whatever you do, don't listen to the Internet. At least not without testing it first.