If the fastest valve action possible along with higher lift, was always better...... then a swap to a higher ratio rocker would always produce more power...... which is doesn’t.
I have done rocker swaps countless times on the dyno.
I’d say the success rate for me on that is under 50%........ with the majority of the time it showed positive results being when the motor was basically “under cammed” or running a “slower” cam.
There can be a lot of things that can impact the effectiveness of really fast valve action.
I have done hardly any dyno testing on motors with tiny cams, but my suspicion is that these are generally under cammed engines, and as such will respond over zealously with added area under the curve.
I wouldn’t expect the same results across the board in race engines that are already running a suitably sized cam for the application(which is the situation for most all the motors I’ve done rocker swaps on), especially if the valve area and flow at peak piston velocity is already sufficient to keep up with demand.