Because public voting is the status quo, so 3 & 4 are identical in what they would (not) change in the procedures. 3 is just an additional statement of support. If you support public voting (3) you also support 4 over 1/2.

Not necessarily. As I understand it, 4 (NOTA) is special, because in a vote to change the constitution of Debian, you need 3 times more votes for a change than NOTA votes. Thus 3 > 2 > 4 would mean "I prefer 3, but if the majority supports 2 I will side with them instead of against them". In this case, I can imagine people thinking "I personally think public voting is fine, but if a lot of people think we should have secret votes, I have no problem with changing the constitution to allow those."

Public voting is not status quo outside of Debian. In the real world the only places that have public voting are places that are controlled by dictators.

It’s very common for boards and other organizational entities to use public voting. I’d argue that it’s a feature of legislative processes.

I do think ranked choice voting is dumb and adds low value and high complexity. This thread with hundreds of comments affirms that notion.

Most (all?) US legislatures don't have private voting and are not controlled by dictators; not sure about other countries. Voice voting does happen in some contexts and doesn't leave a record of who voted how, but most contested votes are recorded.

To me, Debian Developer feels more like a position of responsibility and representation than a member of the public. But I'm not one, and I don't know what the expectations are.

As well as many, many organizations that aren’t governments.