[thaumasiotes]

> In other words, the writing system being a detailed model of phonetics is not necessarily a good thing.

> No romanization scheme captures all the phonetic nuances of Japanese.

The fact that alveolar and palatalized sibilants both exist as contrasting phonemes is not a "nuance". It will be represented in every writing system that anyone ever puts forward, as indeed it already is.

The only advantage of putting 'si' in a Romanization of Japanese is that it corresponds well to the official alphabetical order of Japan. There is no other reason you'd do it.

[kazinator]

Well no; we have hit on that in a private (e.g. one person's own personal) system for breaking down verb conjugation, with the help of roman representation, there is an advantage to using "si" for し, in that they can have have a single stem like "hanas*" that goes to "hanas* -u", "hanas* -anai", "hanas* -imasu" and so on. In that system, they don't have to deal with the phonetic detail that si is alveolarized: which has no bearing on the conjugation logic. The conjugation logic will correctly produce something that contains "si", which is understood as "shi", completely independently/orthogonally.

Romanization systems that use "si" and its ilk are obviously out of favor for the purposes of romanization; I'm not proposing to popularize that. (And really, romanization as such should be largely avoided; relying on it is a trap for learners).

But there is a small advantage in that if you are typing Japanese with romaji-based IME on a device, it is two keystrokes to code し using "si" compared to three keystrokes for "shi", so why not.

I'm not sure what you mean by "exist as contrasting phonemes". There do not exist two phonemes "si" and "shi" at all, in Japanese, let alone as pairs that can be substituted in the same spot of a word to change its meaning. If you speak such that you substitute si for shi, you will still be understood. There are foreign accents like that in Japan. I've also heard unpalatized "ni".