However, the literary (or even plain written) language does not permit this latitude. The Tamil signboard on Dor'saami Road in Madras will only say Duraisaami Saalai. In Bangalore, the common man's Dom'looru is still, on paper, Dommalooru. The integrity of the syllables is maintained in the formal language.
Except in Hindi.
The Hindi newsreader will elide the lagging A's as eagerly as the rickshaw-wallah. The student at a college canteen and the scholar declaiming on a podium would both say, "Yeh to bilkul theek hei." Why is this so?
In my opinion, it's due to the influence of the Persian- and Arabic-speakers who migrated into northern India. And it's not so much due to their language as due to their script.
The Semitic scripts all share an interesting property -- they have no vowels. Their words are represented using only consonants. Such a script is called an Abjad. The word "Hindustan" (e.g.) would be written HNDSTN.
Contrast this with the Indian scripts: We do have vowels. In addition, every consonant character possesses an implicit vowel A. क is not K, but Ka.
Now let's take the word Mehel (as in Taj Mehel) and transcribe it in an Indian script, but with an Abjad spirit:
Mehel --> MHL --> Ma-Ha-La (महल)Or take Neher (canal):
Neher --> NHR --> Na-Ha-Ra (नहर)Or, Matlab (meaning; not the software):
Matlab --> MTLB --> Ma-Ta-La-Ba (मतलब)Such a practice might have started with Arabic/Persian words, and eventually caught on with all words in Hindi.
Because of the Abjad influence, the letters, when at the end of a word, lost their implicit A vowel. And elsewhere, they picked up new sounds such as the short E vowel - as in Mehel or Neher. This vowel sound was not present originally in any north Indian language.
But for other vowel sounds already available in Indian scripts (such as i, e, u) the Persian/Arabic words must have been written in the traditional Indian way - Bi-La-Ku-La (बिलकुल).