It seems to be in the zeitgeist to talk about accuracy before velocity, so I always expect to see these comments in this kind of thread.
From my personal experience, however, I have rarely had to choose one over the other. I’ve found many accuracy and velocity nodes above published book maximum load data, such my maximum accuracy AND maximum velocity went hand in hand. I’ve almost exclusively developed my loads to be near the maximum standard velocity for any given cartridge - because if I didn’t want all of that speed, I would simply use a smaller cartridge which didn’t eat so much powder to hit that speed.
It’s an observation bias which makes so many folks believe the best accuracy is typically found “below maximum”. At the other end, although never considered and certainly never discussed at similar volume, reloaders with equally find the best accuracy is typically “above minimum.” Load data is developed to produce accurate loads within a “window” between min and max. Like a window in your home, you can only see what is visible within the window - so when you plant a flower garden outside of the home, you purposefully center it in the view of the window. Inevitably, we’ve developed load data for generations to produce accurate loads within the pressure window - why are so many folks surprised to find that which data developers planned for them to find? Why are so many folks surprised to find what they seek, where they seek it? And why are folks surprised to not find something in a place they never look?