...Thin Is Sharp.
...Thin your blade out so that it cuts well and keep going thinner until either you are satisfied, or the blade chips or otherwise fails for your cutting tasks. At that point, add a microbevel to strengthen the edge, and you'll have a thin *and* durable edge.
Carter says something similar and adds that almost no one EVER gets to the failure point in his experience. Currently I'm doing some experiments on my Delica in ZDP-189. I've got the primary bevel at something like 12 degrees per side and hope to end up between 8 and 10 degrees per side. It only has a microbevel from being touched up on the SharpMaker. The next sharpening will remove the micro and drop the edge angle another few degrees. So far, the cutting performance is MUCH better *and* it's holding it's edge longer!
Focusing on angles behind the edge-shoulder is, imho, the wrong way to look at things, as stated in the quote above it isn't
angle that cuts or provides strength, it is
thickness.
For most knives,
the primary bevel angle is ~5-dps or lower to allow a thin knife to slide through material with a minimum of wedging. My Spyderco CalyIII in ZDP-189 is ~3-dps in the primary. The hollow-grind of many knife is even lower, the very point of using a hollow-grind. But that is all in the primary and on knives built of relatively thin-stock to begin with. The CalyIII is ~1/8" thick at the spine, FFG at 3-dps across the ~1" wide primary bevel brings the edge to ~0.015" thick at the shoulder, whereupon I sharpen at 15-20 dps to the apex.
The important measurements to consider here for most cutting applications are that sharpening angle and that edge-thickness.
Sharpening at 15-20 dps produces an edge that is 2:1 height vs thickness, an excellent ration for strength as well as cutting efficiency. Everything from hard-use wood-chipper and chainsaw blades to light-use fillet knives can be sharpened at this angle for high performance in a myriad of applications. There is a good reason so many angle-guide systems are set this way.
But BEHIND that edge is where edge-thickness really comes into play. That CalyIII's 0.015" is sufficiently robust for a wide array of tasks, but it is inferior for specific tasks. For example, I have custom-knives and straight-razors ground to <0.005" at the edge-shoulder - 3X thinner and 3X more efficient at penetrating material. An edge that thin is perfect for cutting deep into soft material, even abrasive material. Keep in mind that a standard utility-blade box-cutter is ~0.017" thick. With an edge that thin behind 15-20dps, the edge-bevel is quite thin as well and so VERY easy to re-establish when sharpening, so little material is removed. The CalyIII bevel is more difficult to maintain because it is so much wider, much more metal must be removed to re-apex a dull blade. However, that 0.005" edges are easily bent/twisted if too much lateral force is applied, such as cutting into hardwood or bone - it is not built for "hard-use" but high-performance precision cutting. Remeber, passed that edge-shoulder the knife increases in thickness at only 3-4 dps. Thickness translates to
rigidity (i.e. resistance to lateral stress) in cubic proportions, i.e. when you double the edge-thickness you strengthen it
8-fold. Many outdoor knives are ground to >0.020" thick at the edge-shoulder. Such an edge penetrates relatively poorly but is built to withstand
significant lateral stress, preventing the edge from flexing and potentially suffering a catastrophic chip. Axes and dedicated chopping tools are obviously designed to be even more robust to endure the stresses they must encounter.
What i am trying to get it is that your knife blade geometry should be built for the task it's intended for, and that won't have much to do with a ratio of primary:secondary bevel widths/angles. You start with as thick of stock as you need to provide lateral strength to the blade, and as wide as you need to minimize wedging in the primary bevel or the edge-shoulders without compromising the strength of the blade. If you're experiencing wedging at the edge-shoulders and don't want to re-grind the primary,
then you cut a 'relief bevel' be it V or convex (better). My recommendation is to cut it at ~5dps lower than your apex angle. so if you have a primary grind ~5dps and edge-bevel ~15dps, cut the relief at 10-dps.