Just a reminder: as mentioned many times in this forum, eddy current braked dynos will not show power differences with drivetain weight/drag changes. This includes Mustang dynos and DynoDynamics dynos. Only inertial dynos measure that.
Loss of drivetrain weight doesn't really add power.
Your engine is putting out the same amount as before. This is what an eddy current braked dyno is measuring.
However, when you lose drivetrain weight (or more precisely, drag), your car will accelerate more quickly. An inertial braked dyno is measuring this accelerative force (you're basically turning a big heavy drum).
So does a pulley add power? No, not really. However, does it change your rate of acceleration in the same way adding power does? Yes, definitely.
Lighter wheels/tires have the same issue/effect. They allow you to accelerate faster, but do not add to your engine horsepower.
Your dyno operator should have explained that to you, and he did a disservice to you by not. Dynos are GREAT tools, but they're only that: tools. How they're used and how the information is presented to you really determines what you, as a customer, gets out of it.
I hope this helps!
What you state, specifically about how one dyno can see the rear wheel power difference netted by a UDP while the other cannot, goes against how physics and Newton's 2nd Law operates... Here is why:
Take a crate engine rated for 500hp with zero serpentine driven accessories and throw it on an engine dyno and dyno it. Your results will be pretty close to 500 crank hp as long as the company didn't lie to ya. Now take this same engine and add the serpentine driven accessories such as water pump/clutch fan, alternator, A/C compressor/clutch assembly, PS pump, and a few idler pullies and tensioner. To keep this whole thing simple for example purposes lets say that we just added 5 serpentine accessories that each require 20hp to drive when the engine is at a redline of 6,000 rpm's. That it a total horsepower requirement of 100hp just to turn those serpentine accessories at redline. Put the same engine on the same engine dyno as before and now dyno it. Your power output at the crankshaft will now measure close to 400 because the other 100hp is already being used to turn the accessories and thus is not being used to overcome the resistance to acceleration that the engine dyno is providing.
Think of this like you (1 manpower) while wearing a jacket made of lead pushing a shopping cart full of bricks to a speed of 5mph. Now take the same man and same cart full of bricks but imagine that you ripped one sleeve of that lead jacket off thus lightening your personal load a bit. It will now take less time to push that same cart (aka load) to 5mph because more of your total power can be dedicated to just push the cart rather than carry the full weight of that heavy jacket too. Thus the acceleration is quicker. This is because even though the same amount of total force of 1 manpower is still present in both situations, some of it was already used in just moving forward with the lead jacket let alone accelerating the cart full of bricks to a given speed. Now replace the man with a vehicle engine, the lead jacket with the drag caused by serpentine driven accessories, the ripping the sleeve off with a UDP, and the cart of bricks with the drivetrain of a car and the dyno rollers and you should see the picture more clearly. This is why a UDP technically does not "add" hp per se but it does free up hp that would have been otherwise used to turn serpentine accessories rather than turn the rest of the drivetrain of the vehicle. The less percentage of the total engine power used to turn serpentine accessories means the more of it remains to make it to the rear tires to accelerate the vehicle.
So that engine started out rated at, and making, an honest 500 crank hp. Then we added the serp accessories which reduced output that is usable by the drivetrain to 400hp. Speaking of the drivetrain that also causes parasitic drag too which is roughly 18% depending on many factors. So now by the time we get to the rear wheels and what power is actually remaining to accelerate them we have a total of 328rwhp. Now whether I reduce parasitic drag loss at the serpentine accessories or via the drivetrain this still translates into more power making it to the rear wheels.
Whether you measure this difference on a loaded eddie current dyno or an inertia dyno both will show the difference. The basic principle of the calculations for both loaded and non-loaded dyno's are the same. Force = Mass * Acceleration which is Newtons 2nd law by the way. In the case of the inertia dyno the Mass remains constant; but the Mass does change in the eddie current loaded dyno because the load cell is adding load resistance. This of course will mean that acceleration will drop proportionally when using a loaded dyno. But then again it is Mass and Acceleration that are multiplied by one another. Increase mass by a given percentage but decrease acceleration by that same given percentage and you will still end up with the same force in the end as if you had decreased mass and increased acceleration by that same percentage.
Baseline Example: 1,000 pounds * 1,000 revs/second squared = 1,000,000 lb of force.
Example #2(increase Mass by 20% while decreasing Accel by 20%)
1,200 pounds * 833.33333 revs/sec squared = 1,000,000 lb of force
Example #2(decrease Mass by 20% while increasing Accel by 20%)
833.33333 pounds * 1,200 revs/sec squared = 1,000,000 lb of force
What this translates to is that when you vary the load placed on an engine while it accelerates it also inversely affects the engine speed by the same percentage when at a given force output. This means that whether you use an inertia dyno which has a fixed mass to rotate or an eddie current loaded dyno which can vary the mass your read power output will be the same. This is assuming a perfectly sterile testing environment where both dynos are calibrated perfectly to one another. Most of us know this is not the case in real life, but the numbers do illustrate the point of how a given force can be accurately measured using different dynos. The way Newtons formula operates is also why all dyno roller drums don't have to weigh the exact same to be able to accurately measure power.
The MAIN key to remember in all of this is that whether you are measuring your vehicles rear wheel hp and tq with an Eddie Current loaded dyno or an Inertia dyno it is measuring the amount of hp remaining to accelerate the rear wheels after all other losses due to parasitic drag have occurred. The eddie current loaded dyno will show the same percentage (not amount) of gain in a before and after dyno as the inertia dyno will show with regards to a UDP. The reason I put emphasis on percentage and not amount is due to those real world factors coming into play like dyno operator monkey using a retarded correction factor in the dyno software or other factors that can cause one dyno to read higher than another on average.
Noble please do not take what I posted as any sort of jab at you, your business, or your knowledge because I do not mean it that way at all. All I hear are great things about RPI. But I believe it pertinent to clear some of these misunderstandings up even if I am just a newby on this forum. Most of what was said in your post was dead accurate.
And yes, I am aware I probably over-explained the hell out of this.