Do you think that our speech patterns are so standardized and monotonously the same that people hear and recognize those distinctions in everyday speech? Regular, real people, not linguists or someone making a study?

Most assuredly, yes, but not consciously. That's how we can detect "foreign" accents. For example, most USans can distinguish between the two vocalic sounds in ship and sheep, but some can and others cannot distinguish between cot and caught. An experiment for people for whom the latter pair are homonyms, can you hear the distinction that other regional accents (dialects) make between those two sounds? Whether one can identify those sounds in isolation, or better yet, whether one can disentangle what one hears from what one writes is another matter entirely.

Ceci n'est pas un seing.