Asked as someone learning Japanese. Findings drawn from Japanese-specific research are framed for Japanese learners and may not hold for other languages; where the evidence is general second-language research, the report says so.
Immersion (living in or being surrounded by the language) gives a big boost to learning, especially for pronunciation and speaking fluency, but the evidence here shows it isn't a guaranteed shortcut to full fluency on its own — how much learners actually notice and correct their errors, how much structured feedback and practice they get, and individual differences in things like auditory processing all matter a lot too. None of the sources are specific to Japanese immersion outcomes for grammar or vocabulary, so this answer draws mainly on general second-language research (with one study of Japanese learners of English on the pronunciation side).
Among Japanese learners of English, simply having had immersion experience (time spent abroad) accounted for a large share of how well learners mastered a notoriously difficult English sound contrast, more than classroom instruction alone.6
Even with immersion, learners' natural ability to pick apart fine differences in sound (their 'auditory processing' skill) still predicted how well they learned new pronunciation contrasts — meaning two people with equal immersion time can end up with very different results.6,7
In plain terms
Research on Japanese university learners of English found that speech outcomes were tied not just to time spent immersed abroad, but also to the length of classroom instruction and specific pronunciation training received — suggesting immersion works best combined with active training, not as a substitute for it.7
In studies of French immersion classrooms, students who were simply exposed to the language without targeted correction did not reliably fix grammar errors on their own. Explicit teaching plus feedback that pushed students to notice and self-correct their mistakes ('prompts') worked better than just being corrected for them ('recasts') or receiving no feedback at all.2,3
In plain terms
Example
In immersion classrooms, vocabulary and grammar mistakes were most likely to be fixed on the spot when teachers used feedback that made students actively work out the correction themselves, while pronunciation mistakes were more often fixed simply by the teacher repeating the correct form back.3
In a virtual reality vocabulary study, learners who were less successful at picking up new words benefited significantly more from immersive learning than from simple flashcard-style pairing, while already-strong learners did about equally well either way. This suggests immersion may raise the floor for weaker learners more than it raises the ceiling for strong ones.1
Across many studies, learners who engaged in back-and-forth interaction and negotiation with others (not just passive listening or reading) showed larger and longer-lasting language gains than learners who didn't.8
A study of adult learners of French studying in an immersion program in Quebec found that how difficult learners perceived a speaking task to be — especially difficulty finding the right words — was closely tied to how fluently they actually spoke, showing that immersion outcomes depend heavily on what learners are asked to do with the language, not just on being surrounded by it.4,5
General second-language research indicates that learners who start later in life (after childhood) typically do not reach the same across-the-board native-like mastery as those who start earlier, regardless of how much immersion or exposure they get as adults.9