I really don't understand how such a simple problem can cause such a stir. All this talk about what is best. Isn't it really simple in the end?
The thinnest a blade can physically go, is by going from spin to edge in one staight line, full flat to zero edge. To go thinner than that requires changing spine thickness. Now there are a few provisions to this statement.
1) If you do mainly shallow cutting, you can hollow out the blade, keeping it thin for as long as possible. The extreme of that is a 1/1 hollow ground razor. It doesn't get thinner than that because it employs already a second radius without which it would deform already during shaving. But for through cuts you gain little, because on the top of the blade you run into the spine at an even more sever angle bound to increase drag.
2) If the steel or the application does not permit a full flat to zero edge or you run into binding issues you need to increase the thickness at the very edge while otherwise maintaining the blade as thin as possible, which naturally leads to either full flat with edge bevel or full convex or full flat with convex edge. Which one of these is again irrelevant to some degree. Really
the imperative question is how thick is the blade behind the edge, because much above that we are dealing again with pretty much the same geometry.
2a) Now finally there are are the issues of
drag and binding and while I have ignored them for the longest time I am more an more convinced that they play a bigger role than many think. When batoning (which admittedly is nothing for extremely thin profiles but I think in the end it translates) my experience has been that a convex profile outperforms a flat, most likely because of the smaller contact area (binding). I think even though it took ME a long time to figure this out, this is rather well known and the reason why practically all axes and splitting mauls are convex. If you want to you can create an even bigger stand-off by combining convex edge and hollow grind or applying a median grind, eitherway you have a relief and the blade is contacting only on its widest part (locally of course, draw it up if you don't know what I mean). And finally drag: Any sharp transition creates drag, so obviously you don't want any sharp transitions. Whether you want to believe that it makes a difference or not is up to you, but the difference is there. A smooth transition will create less drag than a sharp transition, which favors the convex grind over the edge bevel grind. Not to mention that the convex grind is locally actually thinner, because for the same angle at the edge and same angle of the body of the blade and same spine thickness, it is create by removing and blending the shoulder -> less material.
Between bevel (not microbevel, they hardly change the geometry) and convex with all other things equal (angles and thicknesses) the convex seems to me clearly superior....by how much? Try it out for yourself, if you don't see a difference, forget about it.
Which one is best between Full flat, convex and hollow and any combination thereof? Take you pick, preferably suited to your intended task, and preferably suited to your choice of steel and heattreat.....
sheeesh is that really so complicated? 
