There are several problems with the interpretation of the opposition of simple and compound adjectives in the ancient Slavic languages as a way to express indefiniteness/definiteness of the noun.
First of all, I can think of no reason (and no typological example) why the definite article would merge with the adjective and disappear as a separate element. Taking into consideration that in the early Slavic texts the structure of compound adjectives was pretty transparent, they seem to have grammaticalized not long before, and this и/ꙗ/ѥ should have preserved (even occasionally) as a noun identifier as well, which is not so in reality.
Second, the meaning of the separate и/ꙗ/ѥ in the old Slavic texts, at least in the Nominative, is "which", thus the meaning of «новоѥ» being "new-which" rather than "new-the".
Third, in Lithuanian, the same opposition of simple/compound adjectives is used (as the grammars tell) to put the emphasis on the adjective rather than the noun: «žalias obuolys» means both "a green apple" and "the green apple", while «žaliasis obuolys» means "a/the apple which is green, which differs by its green color".
Thus, in the opposition «зелено ꙗблъко» vs. «зеленоѥ ꙗблъко» the second variant puts an emphasis on the uniqueness of the meaning expressed by the adjective (the green apple among apples of other color). With time, the meaning shifted: in some languages (South Slavic ones), towards expressing indefiniteness/definiteness of the noun; in others, the meaning became identical, with differences in syntactical usage or in stylistic flavor.
By the way, you can find this across the literature and even in manuals of the Old Slavonic. Simply, since this distinction seems to be absent in any modern European language but Lithuanian, people by default don't pay attention to the subtleties and classify the old definiteness of adjectives as just a special case of definiteness of nouns.