If the pi actually is being used it'll definitely draw 2.5-3A, so a 2Ah battery won't last very long. I'm not really up to date on li-poly (the flat pack lithium style batteries), perhaps there's some that would be ok for a handheld with ~3A or more of draw, but again, it's another $20-$40 for batteries that are usable for a handheld. There is the pi arcade hAT or whatever, that's about $15+ iirc. 3d print a case, if you already have a printer that's just time and materials, otherwise you're gunna pay $20 for a case design and print that actually feels nice, again.
At a certain point, the steam handheld device might be a better bet, since it'll have a warranty (don't laugh!) and a support team (okay laugh a little).
All of this could be irrelevant because perhaps the pi zero does everything people want, and that's dirt cheap and also probably uses way less power.
LM2596 costs <$1 and it powers my rtl-adsb for 4 years so far (yes, this is in step-down mode, but you either want a 2S/3S battery anyway, or you use the similar step-up chip).
I have a bunch of "QC3.0/3A Buck converters" for 6-36VDC to USB-A, maybe i should try to run a pi on a SLA Gel 12V battery.
> While this is admirable as a proof of concept, I suspect that moving the price point up to, say, $30 could give a much more useful and hacker-friendly device.
I agree. I think this project proves that ultra low cost is viable. Now relaxing the constraints a bit to add RAM and WiFi could iterate it right into an optimal spot.
I wouldn't want to go much farther than RAM and WiFi, though. It's too easy to let these projects turn into a slippery slope.