Let me chime in here. I've read the explanations above and the one I agree with the most is @tor47 . I've read and understand the other explanations, but having the valve stay open too long and dump excessive air into the transfer ports and barrel wouldn't slow the pellet down IMHO. How would that work? The pellet has already left with a specific amount of air, and then as dwell increases its putting
more air behind it. That could affect accuracy by having turbulent air behind the pellet as it exits the barrel, but I'm not seeing how that would slow down the pellet or cause it to be traveling at a slower velocity.
It
does make sense to me that as HST increases to a certain point, we get to the "plateau" of the HST/velocity curve, since past the knee and onto the plateau, further spring pressure doesn't open the valve any further or any longer. But to keep going until the spring tension causes the hammer to fully compress the valve spring to the point that it causes the valve to "bounce" at the end of its travel and start closing faster than it normally does would cause the valve to shut sooner than with less spring tension and result in less dwell and a lowering of velocity. My two cents...