My personal optimization is also to not make noise at all, if I can clearly and safely pass without warning i.e. the pedestrian is not moving erratically and there are no small children. On rare occasion this elicits anger from (usually elderly) pedestrians though.
For some humans, sure. The trouble is that maybe 1/3 of people simply won't process the "ding" at all, which makes it so unreliable as to be useless in practice, IME. (You can ring repeatedly and people eventually notice, but overall that's less effective than saying something).
How so? You can say something in the same scenario where you'd ring the bell.
> It also can be heard earlier than a biker's speech (assuming the biker doesn't continually give verbal warnings for 15 seconds as is true for basically all bikers on earth).
I don't quite understand the point you're making here - most bikers don't continually ring a bell for 15 seconds except if a pedestrian is obviously not paying attention, and speech is more effective in that case. Maybe a bell is louder than typical speech, but since it's not acted on reliably by pedestrians, that doesn't help much - you always have to be prepared for the worst case.