Proving that too sharp hooks don't work when they don't lift off: easy.
recurves that don't lift off add mass to the bow. Mass that's not working at any point during the draw or increasing leverage. So you might as well have a shorter recurve and add lead to them. The longer it takes for a recurve to lift off during the draw, the longer it's dead mass. Sure, it gives a fatter force-draw curve. That's expended energy. Large part of that fatness is needed during release to move the heavy tips instead of the arrow. Hence, you're losing efficiency.
See it in terms of leverage and tip mass. Recurves have high early draw weight because early in the draw they are functionally shorter bows. A 60" bow with 5" hooks is at brace a 50" bow with the mass of a 60" bow. The functionally shorter bow strains the wood more at each increment of draw, so you have higher early draw weight. Later during the draw, when the string lifts off, the extra leverage comes into play. The bow becomes functionally longer, with more leverage, so less draw weight gains per draw length increment. But only when the hooks are cantered 90° relative to their original position is the leverage fully realized. Before that, the bow is still functionally shorter than its initial 60". But at all times, it carries at its tips the mass of the 60" bow. So the trick is to find where this trade-off between extra mass and extra leverage is optimized.
Kooi & Bergman 1997
http://www.bio.vu.nl/thb/users/kooi/kobe97.pdfshowcase this very nicely. With the extreme Hickman working recurve (2/3 of recurve) as a good example to contrast to a normal working recurve and static turkish and persian recurves.
https://www.flickr.com/photos/farflinger/5177069236/in/album-72157625386947492/ This bow design was patented by Hickman as a revolutionary new bow design. String lifts off completely at about 21". That means that before that, it's functionally a shorter bow, but with the extra mass of a longer bow, concentrated at the tips. The reason we aren't all shooting this high-energy storage bow is that it's not efficient at all. The fat force draw curve deceives the eye but not the arrow. And neither the mathematic model by Kooi & Bergman. In fact it isn't faster than a well designed straight stave bow with a flat force draw curve. Anything above the straight line in the force-draw curve of that bow is energy wasted in moving a returning limb with a portion of dead mass that becomes larger during the return of the limb. With cartilage-unfriendly hand shock.
If you don't buy this explanation: experience for yourself. Make a recurve with 90° hooks that don't lift off. And make them looooong. 10". And make a recurve with shorter hooks, that lift off early during the draw. See which shoots best. If you don't want to waste good bow wood for this, make these example bows out of PVC. Will take you 1 h in total.
it's just a matter of dead mass caused by bow parts that cannot store nor impart energy to the arrow by either working or adding leverage during the entire draw.
There's a reason why the fastest @10gpp wood-only bow built so far is a long straight stave bow without sharp recurves...
Joachim