Practical how? It's literally an order of magnitude easier to implement Korean or Chinese TTS than English. A Korean block or Chinese character only has one pro

Practical how? It's literally an order of magnitude easier to implement Korean or Chinese TTS than English. A Korean block or Chinese character only has one pronounciation, with the number exceptions few enough to list. The mapping from symbol to sound is near deterministic. English TTS has been under development for decades and it usually still doesn't pronounce half the words right. They also don't require the convoluted kerning and line breaking algorithm that English text needs. Unlike for latin, Monospace spacing is canonical, so you don't have to do the weird dance where code has to use one font and prose another font. Literally the only argument you can make is that if you mutilate English spelling enough you can get do a 8 bit per character encoding, but that's a dumb point because a 16bit Korean or Chinese character usually represents more information than 5 English letters.