Reading out-loud is a skill that needs to be developed, but I do not believe that it leads to internalization or fluency of the language.
Would reading German stories out-loud lead to German fluency? Reading out-loud can even slow down comprehension. It's a 'seminary legend' that it helps internalization.
So what does "reading out-loud" do? Is it helpful? I would say absolutely YES. Especially YES, if aural comprehension is happening when you read and hear the words you speak. But if a person is not at that stage of aural comprehension, reading out loud is still one of the skills that is needed. One needs to take into account what happens when one reads. Is following a text while listening to audio of value? Is following a text while shadowing it (reading it out loud while listening to it) of value? (I recommend every read Grabe's book Reading in a Second Language
So what happens when a person reads?
(1) When a person reads a text, they first convert the written text into audio in their working memory. (This is a well established fact). This means that the oral/aural component is a primary component of a successful reader. (You can Google 'the phonological loop' to learn more about this).
(2) Some very common words are perhaps not vocalized when a person reads, e.g. "a, the, is, on, are" etc. (This collection of words, as it increases in a person's 'sight-lexicon' is another basic element of a successful reader. Second language reading experts say that a person needs to be able to read 180 words per minute in order to have comprehension [remembrance and the ability to analyze and answer questions about a text]. The most frequent words get turned into sight words.
(3) A lexicon of phonological expectations (the allowable vowel/consonant clusters) needs to be developed, which is then accessible to the working memory, so that syllables, word roots, endings, etc. are easily deciphered, and word/morpheme boundaries become automatically deciphered. This only happens by both hearing and producing those vocalic/consonant clusters that get interpreted as syllable and morphological components. Reading out-loud is one way of building one's phonological lexicon of expected/allowable consonant/vowel combinations. (One never would hear or read φιφ, θιθ, χιχ, κτλ).
(4) Vocabulary is retrieved very quickly, and it is done via an audio lookup in a person's mind. If a reader does not have the wide spectrum of a well-developed phonological lexicon, they will struggle with vocab lookup.
(5) The working memory then builds up a Text-model (a rather literal understanding of the text without much appeal to interpreting the text via understanding the greater context (the situation model).
(6) The working memory is limited. So it can only process so much audiolingual analysis, form-lookup, and comprehension of what the text says, at once. If all resources are put towards form-lookup or syllable analysis, few resources are left for comprehension, dis-disambiguating multiple meanings, ktl. And in such a resource intensive demand, comprehension fails.
(7) A fuller understanding of the text is developed as one draws on previous understanding, the genre, expectations, etc.
But at the core, throughout all of a person's effort in reading, is the presence of aural/oral components. Those students who have learned by the communicative method (speaking/listing to someone and trying to understand what they are saying, negotiating meaning, verbally asking questions or asked questions) have developed their phonological lexicon extensively. They have also built their vocabulary more than the traditional student and are more able to access that vocabulary because of their aural/oral preparation.
The Grammar-Translation student, who has learned mostly by reading silently, has not developed a strong repertoire of evidenced phonological clusters. Thus they have a more difficult time reading, because their audio repertoire is often deficient. They may not have the ability to separate prefixes from the root, e.g. συμβαίνω > συν-βαινω (some words are much more difficult). Their ability to look up words with any degree of quickness is slower, because they have not developed their audio lexicon. Reading out-loud is one way of developing such a repertoire. Reading out-loud makes a person decide where syllable breaks are and then incorporate them into their phonological lexicon. Reading out-load adds audio to the otherwise silent learner's 'silent' method.
Without going into more detail, there are models of second language acquisition which are based on "comprehensible input" and "comprehensible output." What comprehensible
means is the ability to understand because of self-apparent content of the L2 phrase and the person's background knowledge and skill level. Output means via speaking
; Input means via listening/hearing
. Modern language courses such as Assimil incorporate both the aural and the oral into their exercises. Whether they do this because they believe in comprehensible input or comprehensible output is not always apparent. But learning a language requires four different sets of skills: listening, speaking, reading, writing (Some of us call these elements by the acronyn γάλα - γράφειν, ἀναγινώσκειν, λαλεῖν, ἀκούειν). The first two skills are based on the last two skills).
Learning a language is a gestalt experience. Using only one or two of the avenues of learning means that a lop-sided model of second language acquisition is being utilized. Reading out-loud is not an act of communication. And to that extent, it will not help you learn the language in and of itself. But it helps develop the needed aural/oral skill-set that is needed for communication and comprehension to occur.