Dialectal variety is true for many other languages, though, which nevertheless manage to do some kind of standardized spelling.
Thing is, in most cases when it's claimed that "spelling reflects pronunciation" for some language, it's not actually true - what it reflects is the phonemes of the language, not phones (actual sounds). To give an example in English, /t/ is a single phoneme, but it can be realized in a bunch of different ways ranging from actual [t] in words like "tea" to a glottal stop [ʔ] in words like "butter" (frequent in UK). Native speakers often don't even notice that these are two completely different sounds, because they are mentally mapped to the same underlying phoneme.
But because our brains are already perfectly capable of handling such mappings (so long as they happen according to consistent rules, which they almost always do), a phonemic spelling works great in practice. And when you deal with phonemes, the difference between various accents and dialects is usually much less than when you look at the phones.
English is problematic because its spelling is strongly inconsistent even with the phonemes, never mind the phones. And, yes, the rather extreme dialectal variety of English means that different dialects do have some phonemic variety as well. But, again, it's much less than actual pronunciation differences would make you believe, and it's certainly possible to come up with a canonical mostly-phonemic spelling that would cover, at the very least, British RP and American GA, and be reasonably adequate for most offshoots of those.
It's also possible to just adopt phonemic spelling based on specific dialects and expect the speakers to handle the differences at least for the popular ones (just like we already do with different words like "elevator" vs "lift"). This was the approach taken by Serbo-Croatian, where e.g. "vreme", "vrime", and "vrijeme" would all be valid spellings of the same word.
Thing is, in most cases when it's claimed that "spelling reflects pronunciation" for some language, it's not actually true - what it reflects is the phonemes of the language, not phones (actual sounds). To give an example in English, /t/ is a single phoneme, but it can be realized in a bunch of different ways ranging from actual [t] in words like "tea" to a glottal stop [ʔ] in words like "butter" (frequent in UK). Native speakers often don't even notice that these are two completely different sounds, because they are mentally mapped to the same underlying phoneme.
But because our brains are already perfectly capable of handling such mappings (so long as they happen according to consistent rules, which they almost always do), a phonemic spelling works great in practice. And when you deal with phonemes, the difference between various accents and dialects is usually much less than when you look at the phones.
English is problematic because its spelling is strongly inconsistent even with the phonemes, never mind the phones. And, yes, the rather extreme dialectal variety of English means that different dialects do have some phonemic variety as well. But, again, it's much less than actual pronunciation differences would make you believe, and it's certainly possible to come up with a canonical mostly-phonemic spelling that would cover, at the very least, British RP and American GA, and be reasonably adequate for most offshoots of those.
It's also possible to just adopt phonemic spelling based on specific dialects and expect the speakers to handle the differences at least for the popular ones (just like we already do with different words like "elevator" vs "lift"). This was the approach taken by Serbo-Croatian, where e.g. "vreme", "vrime", and "vrijeme" would all be valid spellings of the same word.