If you pronounced the word “suv”, you would use the ‘a’, but SUV is pronounced “ess u vee”, it’s essentially three separate words and ess starts with a vowel.
It’s confusing until you think about it a little.
But this is written language, not spoken language…
Nevertheless, you are right, it’s not so confusing, (Except that I never really learned the names of the letters in English, but that’s on me).
The reason to use “an” is because it’s awkward to pronounce two sequential vowel sounds. We have a tendency to mash them together. So “a apple” would become something like “ayapple”, and then it’s difficult to understand what’s being said.
So, we insert a consonant in between. It becomes “an apple”, with the “n” consonant making a clear phonetic break between the words. That break between vowels is hard-coded into English. It’s not an arbitrary rule, but a completely necessary one that I believe all native English speakers instinctively understand.
We always have consonant sounds between vowel syllables, whether that’s reflected in writing or not. We just don’t have consecutive vowel syllables. Even words written that way are never pronounced that way. For example, the word “beauty” has three consecutive vowels but they’re all one syllable; in standard American English pronunciation, it sounds like “byoo-tee”. The “eau” becomes a “yoo” sound, as we blend them in together.
So yes, “a” or “an” depends on whether or not the beginning of the following word starts with a vowel or consonant sound, and ignores what actual letter the word starts with.
A curious thing about “an” is how it became rebracketed in the past. It used to be ‘a napron’ before it morphed into ‘an apron’, and then inversely ‘an ickname’ became ‘a nickname’.
While the “n” makes it easier to understand in speech, moving the “n” over to the “a” to form “an” makes it easier to understand in writing. But that’s a very interesting piece of history.