On Sat, Feb 8, 2020 at 2:01 PM Bohdan Khomutskyi <bkhomuts(a)redhat.com>
I'd suggest not selecting any specific area to concentrate on,
achieve an all-good solution that will combine the benefits, without
specializing in one particular area.
Even if you can achieve some modest improvement in all areas together, you
still need to decide whether you prefer "50% improvement in area 1, and 2%
improvement in area 2 and 3" or "5% improvement in area 1, 2 and 3"
numbers used). So you need to decide your priorities, and usually not in
absolute scale ("area 1 is the top-most priority at all costs") but in
relative scale ("I'm willing to sacrifice that much in area 2 in order to
improve area 1 by this much"). Of course this is hard to codify, unless you
fancy writing a science paper with complicated equations, but mostly it's
enough to look at the numbers and pick the candidate that feels like
satisfying your preferences in the best way.