A script borrowed for a language it was not designed for either fails or is altered, and the Tibetan alphabet is a record of the alterations that were judged necessary in the seventh century and afterwards.
Some of the work was subtraction. The voiced aspirates of the Indic order were dropped because Tibetan does not have them, and so was the whole retroflex line, which is why thirty letters stand where the inherited arrangement would give more.
The rest was addition, and it took two forms. Inside the recitation, the affricates tsa, tsha and dza were built from the palatal letters with an added mark, and letters were provided for zha, za and the weak small a, none of which the inherited order supplies in that position. Outside the recitation, an apparatus was assembled for writing Sanskrit exactly: reversed forms for the retroflexes, subjoined ha for the voiced aspirates, and devices for the long vowels Tibetan does not distinguish.
The division between the two is deliberate and has held for thirteen centuries. What Tibetan needs for Tibetan is in the alphabet a child recites. What Tibetan needs for Sanskrit sits beside the alphabet and is taught separately. A page can therefore be read at a glance for whether it is handling native vocabulary or transliterated Indic material, which is a genuine convenience in a literature built largely on translation.