Switch has no games and none on the horizon but assuming it did have something I cared for, or I saw any real potential in it as a homebrew system.
Usual balance of convenience, cost and capabilities. I will also keep in mind that there are millions of theoretically hackable and unpatchable systems out there compared to something like the 360 where there was a sharply limited supply of JTAG capable machines, much less JTAG capable machines unlikely to experience RROD any time soon.
Having a nicely worked up firmware mod during the period the softmod/simpler mod set are busy doing not a lot or infighting (see also most consoles in history where such things exist) is something I could stand to pay for. Some kind of good future proofing or "my little brother updated" protection would be sweet too.
A "nicely worked up firmware mod" would mean simple SD and possibly USB launching, ROM hacking and homebrew launching as well as nerfing things like region checks. Cheats would be nice but if I can do that via ROM hacking then I can live with that.
Similarly I can solder, pretty well for that matter, so that is not a concern but I would rather not spend 6 hours messing with enamel coated wire and debugging strange capacitive/induction/coupling effects or matching impedances.
Cost wise then price of a game or two would likely be the limit.
Is it likely to leave me in the lurch in the future if I have to remove it and go softmod? Will I have to painstakingly reconstruct keys (I would hope I backed anything like that up but I have been known to play it fast and loose in the past) to create a working nand image because updates and efuses sort of thing?
Online might factor into the calculation but I doubt it, even more so if I am supposed to pay for it (some might wonder why I will happy pay TX but not Nintendo, simple answer is TX would deserve my money but Nintendo have done nothing for it).