Should initials() skip Lao, Khmer and Tibetan, as it skips Thai and Burmese? (ສົມຊາຍ ໃຈດີ gives 'ສ. ໃ.') #1380

#1256 gave Thai and the Myanmar script a row in the script-property table saying they write no initials. Lao, Khmer and Tibetan have no row, so they read as Latin does: initials() takes the first character of each word, a leading vowel sign included.

Questions

• Does each script write initials? decisions/script-properties holds the recipe: count CLDR's locale-authored name patterns that produce an initial (lo, km, bo), after the copy test that caught Burmese's block being English's.
• Where one does not, it joins the no-initials column, with the same consequences Thai accepted: empty initials(), a letter with a period not read as an initial, and the name match reading a single letter as a whole word.
• Does each write period abbreviations? Lao titles such as ທ່ານ are spelled out; Khmer and Tibetan are not known.
• Tibetan separates syllables with the tsheg (་), not a space. Is a "word" there what initials() should count at all?