Not that I trust your sincerity, but to answer the question anyway, as I explained in the first post, the resulting settings are based on rough measurements. The batch files take emulated system speed as input, or uses the default, and multiplies by the rough measurement of the given system. DOSBox cycles are equivalent to an emulated system regardless of host. Host speed only determines how much load on the host system is required to attain a given DOSBox setting, and if not fast enough will fall short of achieving the set DOSBox speed. The batch files do set /a cycles=%dosboxcycles%*%cpucycles%. dosboxcycles being the rough measurement of how many dosbox cycles there are to one emulated xt/at/386/486/586 cycle, and cpucycles the desired emulated machine cycle setting. As for rough, the estimates could possibly be made more accurate with some more testing, but they will never be truly accurate since DOSBox doesn't count cycles on an opcode basis, so some are faster/slower compared to a given emulated machine even when set to that level overall, and things are really out of whack when it comes to FPU, DOSBox is running the FPU way faster. Ironically?, on XT machines, the FPU ran slower than the CPU. So, this is as good as it gets until someone adds cycle counting per opcode to DOSBox.
As for your I see no point of the batch files. One, it's much easier and faster to type and remember "xt" or "xt 8" than "cycles fixed 277," and two, setting by system has context from personal association and game options, like "tandy" or "ps/2" or "mcga"(why I added the reference info to the output.)