These are not small things - not only does it take a lot of resources (the stretch goal examples, that is, but the pledge goal stuff can snowball when you have to multiply it by the pledges),
Yes, the things I suggested
are small things. If you already have a system designed that supports party characters and NPCs, then you do not need to change any system to simply add another character the player can choose and interact with. On the contrary, it is a stretch goal that you can calculate pretty well (because the work to do is pretty much the same as for already existing ones, which you have already calculated).
Re Linux/Mac support: It's a good guy thing to do, but it's hardly ever profitable. It really is too bad that so few games are developed for Linux, but a game studio has to spare a thought or two for survival, lest they land in a position where they can have no regard for anything else. Planning for the next game has to be a part of plans for this game.
I cannot agree with this at all.
Personally, I almost never buy indie games that do not support linux, if I do not want them
very badly.
Especially as an indie developer, where you owe your existence to open source and free software (and that in 99% of cases means linux capable software), you have an obligation to support the platform.
Which can be done donating, of course, but a far more logical move is to contribute to the growth of the platform.
And depending on the tech you choose, this can be any amount of work from very little (if you pick an engine that already is cross-platform) to just too much (like, if you for some reason decided to write your engine yourself using only DirectX and Windows API or picked some single platform engine for whatever weird reason).
If your tech basically already supports more platforms, making sure it also works on another is pretty easy and done in few weeks (and involves only the coding part), which suits a stretch goal fairly well.
And if you cannot spend a few weeks because your budget is
that tight... you shouldn't be doing the project in the first place.
And I say that as someone who has already worked on many, many cross-platform projects and porting from platform A to B.
It isn't anywhere near as complicated as some people make it sound,
if you put some thought into it beforehand.
Besides, working on linux is heavan as a programmer, while working on Windows is... way less productive at best.
So it actually makes far more sense to work on linux for the most part, and just from time to time check nothing has broken on Windows, even if 95% of your sales are on Windows.