Regarding the character ROM, this can be found at Bob Zimmers' great Commodore 8-bit archive, see: https://www.zimmers.net/anonftp/pub/cbm/firmware/computers/p...

Regarding kana/kanji – well, my terminology may be fuzzy. But the emulator has "kanji" in the menu…

Regarding katakana vs hiragana: I guess, the simpler strokes were one of the reasons. Moreover, it seems that in daily use it can be distilled down to more or less the size of the Western alphabet. While there are 46 syllabograms in use, 8-bit computers managed to get away with about half of this. (The Sharp MZ-80 is another example.)

For fun, have a look at my own attempt at squeezing hiragana into an 8-bit character generator: https://www.masswerk.at/char8/#U3040 (A rendering demo can be found here: https://www.masswerk.at/rterm/ )