To pick one: the compression of a closed quick release does change the adjustment of hub bearings, easily demonstrated. Various techniques exist to correct the hub bearings adjustment for this. Barnett publishes a technique, a bit fussy, but it works. The technique he demonstrated to me was flawed.  I show him the flaw in logic and he dismissed it. He put pressure on the end of an axle on one side of a hub and pressure on the lock nut on the other side of the hub. There is never any pressure on the end of a skewered axle. The clamping force is only on the lock nuts against the dropout surfaces. His test had two variables instead of one. Does the axle compress under quick release loads or do the threads flex. One test will not answer two variables. Two test are needed to test two variables. Yet only one test is valid. Since a QR can never press against the end of an axle and still have clamping force on a  lock nut against the dropout, pressing against the end of the axle test is invalid.  John' s testing wanted to find a specific amount that a hub could be adjusted from the standard bench adjustment to the on the bike correction for quick release forces. He came to realize that the variables in axle materials and quick release materials and construction were to great for a unified answer. I solved for his issues and his beliefs dismissed my solution.  I demonstrated a perfect adjustment for the added load of a quick release in less than 30 seconds. Once I did six adjustment in 10 seconds.  My special tools cost  $6 from a frame builder.    Special tools do exist, but unnecessary. Most mechanics learn in school or from someone like me, People are asking me for free instruction, is your knowledge free? if not why are you advertising here also? and most develop their preferred method.

 

Chains don't stretch. At 2000 pound of force chains do stretch, True: on a bike they only wear. Cables do. At 550 pounds of force derailleur cables stretch, but the plastic shift lever mount will fail at 200 pounds of force, at 800 pounds of force brake cables stretch. sadly a rim side wall brake surface will start to fail at 160 pounds of force. Numbers don't lie and are verifiable. (Very different structures.)

 

Spokes remain elastic. Spoke elbows deform from the original 90 degrees to the best pull angle, effectively making spokes longer once. On aluminum hubs the elbow indents the hub flange hole making the spoke effectively longer once. Neither of these changes the elastic properties of the spoke, but they do make the spoke effectively longer once and drop the tension of the wheel, allowing for greater tension ranges in spokes, creating early fatigue and spoke failure. Also reducing the efficiency of the wheel to accelerate as looser spokes need to be tensioned by the force from the chain before they move the cyclist forward,  and the wheel has more side to side movement with looser spokes making control and braking less than optimal. 

 

May I suggest familiarizing yourself with current best practice (rather than the noise that gets written on the web).

 

I suggest Sharp for the basic mechanical engineering My experiments with the wheel in both practice and with finite element analysis and  computational fluid dynamic program ANAYS. have shown insight intto the wheel and how to do wheel work 20 faster for better result.   and Barnett for current mechanic best practice. As a BBI certified master tech, I can point out many flaws to Barnett's best practices. From making mechanics less profitable with slow procedures that are not valid in their testing. To just plain bad science understanding.   

 

Get back to me if you have questions after having thoroughly understood those.

 I am back, Now are you ready to learn? or do you have any ideas of your own that are brilliant. I am all ears