Sure there could be a point in doing so. The short anwser with examples: see TBB4 chapter on heat treatment by Marc St-Louis.
The principle is very simple: a certain amount of wood can only take a certain amount of strain (and store a fixed amount of energy) before failure (breakage or bow breakdown due to set). Setback, reflex or recurves will increase total tip travel from unstrung to full draw and therefore strain the wood more, all else being equal. In a straight bow fully drawn, the tips travel 2” less than in the same bow with recurves 2” in front of the handle. Tip travel length is typically around half of the full draw. So 2” of reflex will mean a tip travel of 16” instead of 14”, hence the wood is strained about 12% more. If the wood can safely take that extra strain, it makes sense to flip the tips.
Flipping tips raises draw weight. Not only early draw weight, but draw weight throughout the entire draw. It may be that you end up with flipped tips, and the same final draw weight. This means you damaged belly wood and caused (a lot of) extra set, and the net effect is close to zero.
The fyl in the ointment: If the wood can take that extra strain, it means that your original bow design was suboptimal: you have a bow taking 28” draws, but which can actually easily take 10% more, up to 31”. Remember, the total amount of energy that can be stored remains the same for the same amount of wood.
How to know if the wood can take that extra strain? There are mathematical ways of guesstimating it but few like to do that. Draw it to 30", and see if it explodes, or takes extra set. If it does, you screwed up your bow. Too bad.
A safer way is to use Badger’s no set tillering method. Go slowly up the tillering tree from 25” with increments of 1” and monitor how draw weight changes as you go up and down with 1" increments. If you go from 27" to 28" and back to 27", and draw weight going down at 27" is lower than going up from 26", it means your wood is taking (reversible) set. If it stays the same, increase draw length and repeat process. If you want a 28" draw with 2" flipped tips, you should be able to go to 30".
On a straight bow already strained maximally at full draw, flipping tips is possible, but at the cost of having to thin the belly so as to avoid overstraining the wood at that same draw length. Cast wouldn’t be better, as you aren’t changing max energy storage. As mentioned by others, at that bow length the string angle at the tips changes too little to affect stack (and bow efficiency).
I did cut a few corners, but I think this boils down to the essence.
And for those focusing on string angle: there are two string angles to take into account: at the tip and at the arrow. see
http://paleoplanet69529.yuku.com/topic/21787#.VXfv0vntlBc