Simple: 28% is easier to get up and running. There are a lot of variables here. For my reactors, I use 1/16" FDA grade Buna-N rubber and 5 bipolar 2 active media blasted and specially conditioned 316L 18/22 gauge plates. This allows me to achieve roughly 1 L/m @13.8V 10-12A at room temperature. This is no easy task at all. Lots of work to build a single reactor, but I don't need 28% KOH - only 1-2% by weight. Others are fine with using 28% KOH and getting roughly 1L/m @13.8V >=15A at room temp. Every one of my tests have proven that higher concentrations means shorter bubbler fluid/media life and greater potential risk of engine ingesting noticeable amounts of vaporized electrolyte. My motor costs too much to take unnecessary risks that I can avoid. Also, since my latest discovery in plate conditioning, I have noticed two very interesting facts: system fluid stays clear and gas flame once ignited takes the form of what appears to be a "firework" instead of the traditional flame. I'm still compiling information and researching what may be happening here..