Because this is how the language works, and how it has worked from the very beginning. After decades, only a minority of speakers use li differently, if any.
The particle li introduces a verb to a subject more complicated than only mi or only sina. It's commonly said that li is "omitted" after mi or sina, or that there is a "hidden" li. This could mean that Toki Pona likes to concentrate on mi and sina being more immediate subjects. They hold such importance in the language that anything else is more complex or more external to an immediate experience so that it needs an extra separation by a particle.
Etymologically speaking, li is a third-person marker. It is derived from the Esperanto third-person singular pronoun li[1] and functions similarly to Tok Pisin's particle i, which introduces the verb except when the subject is the singular first or second person pronoun[2][3][4]
In terms of practical benefits, omitting li after mi makes it less likely to be confused with ni. It also makes sentences shorter and slightly more varied in terms of word distribution.
↑Word Origins. Archived from the original on 2 November 2019. Toki Pona.
↑Cite error: Invalid <ref> tag; no text was provided for refs named franklin1980
↑Cite error: Invalid <ref> tag; no text was provided for refs named tung2014
↑Cite error: Invalid <ref> tag; no text was provided for refs named verhaar1991