Maybe pictures of a real life velocity graph of my car on track will make it clearer on the effect of an upshift on real life flat out acceleration.
If you have a look at the telemetry I posted in this older link, see in the first post (and click again to expand the telemetry graph for more detail).
The top graph is a velocity Vs distance graph, of my car on two seperate track days in Mondello International circuit. In general my gearchanges were at 7200, unless the very light track traffic dictated otherwise.
The original purpose of the graph was to compare the effect of DS2500 pads Vs Performance Friction Z-rated Carbon Metallic, but thats not really relevent. The red plot is simply the PErf friction car, and black is the DS2500 car.
Anyway what I would like to draw your attention to is the shape of the upward velocity curve on gearchanges.
The gearchange points should I believe be self-evident
In general at every point when up-changing you can immediately see the effect that it has on the cars acceleration, since the upward angle of the velocity line is immediately slightly flattened.
With a small bit of intuition you should be easily able to sense the shape that the velocity graph for the higher gear would have taken if the upshift occurred much earlier (i.e. at peak power).
I think it is very clear that the velocity graph would still be at a much shallower angle than for revving out the lower gear.
There was at least one or two overtakes there which might have caused a transient lift of the throttle so look for the overall pattern, it still seems reasonably clear I think!
Somewhere I have stored away an overlaid graph of acceleration in 2nd, 3rd, and 4th which is where I measured the effect of gearing but really it just confirmed the obvious - go to the red-line.
End of story?