First, here is an important difference between Latin and English syntax. In English, a preposition often combines with a verb to change its meaning. For example, "I picked up the Christmas-tree ornament" means that I bought it or otherwise took possession of it. "I picked out the Christmas-tree ornament" means that I selected the ornament from among many, after careful consideration. "I picked a Christmas ornament" means that I selected one, without implying thoughtful rejection of many alternatives.
A common way to analyze English phrases like "pick up" and "pick out" is to say that the phrase as a whole has a meaning, which cannot be derived from the individual words. Hence "pick up" and "pick out" are called phrasal verbs. The phrase is the indivisible unit of meaning rather than the individual words. That makes the theory neat and pristine, but it's not the whole truth. More properly,* these phrases make analogies with other phrases in complex and messy ways. The preposition is often chosen to eliminate an ambiguity or to block an unwanted interpretation. For example, "I picked a Christmas-tree ornament", it would mean only that I selected one, not necessarily with thoughtful rejection of alternatives. Adding "out" evokes the core meaning of "out", suggesting emerging "out from" something else. It also makes an analogy with other "verb-out" phrases, as in "I worked out a solution", "His car spun out on the ice", "I need to clean out my desk", where "out" indicates thoroughness and completion, which the verb alone would not suggest.
In Latin, as in the Romance languages, prepositions usually don't work that way. Each preposition has a meaning of its own, independent of the rest of the sentence. In Latin, "in" just means "in".
However, Latin does commonly use prepositions as prefixes on verbs to achieve similar effects. Consisto is itself an example. Sisto means "I stand", and has senses including appearing ("standing") in court. Con- is the prefix form of the preposition cum, meaning "together". In prefixes, the "together" meaning is often extended to make con- an intensifier. People call upon the intensifier meaning when they say:
Consisti in horto. (I stopped and lingered a while in the garden).
If you meant that you stood a while in court, but not as a defendant or witness, "Consistam in basilica" might help rule out the courtroom interpretation (this example is only a guess).
Ovid calls upon both the "together" and intensifier meanings of con- when he says:
ter frigore constitit Hister (the Hister river froze three times)
literally "thrice by cold the Hister 'stood-together' [or 'set']." This sense where "standing together" = "congealing together" will reappear in what follows.
It turns out, though, that consisto does get its meaning clarified or modified by prepositions in a somewhat English-like way—not enough to make linguists say that Latin has phrasal verbs, but enough that you can see the similarity. Lucretius famously says:
Omnis per se natura duabus constitit in rebus; nam corpora sunt et inane. (All nature was/is made of two things; namely, they are bodies and the void.)
The preposition in helps clarify that the kind of "standing together" being talked about here is being made of certain elements or parts. This is a reasonable way to exploit the broad, ordinary meaning of in in Latin, but consistere in is something of a fixed phrase for this meaning. The preposition ex, as in "Societas constitit e pluribus hominibus" (society was formed out of many people), combines with consistere to produce a meaning similar to English "made out of" or "made from".
In English, "Milk consists in vitamin B12" doesn't mean that milk contains B12. In fact, as I read that sentence, I can't tell what it means. Here's why.
"A consists of B" means that A is made of B throughout—just like consistere in in Latin. The Latin sense of "they are standing stably together, forming A" seems to live on.
"A consists in B" is quite vague in English. It usually means that B is the essence or most important aspect of A. I think in here is best understood primarily as a way to suggest a different interpretation than of would suggest, where the details depend heavily on context. To illustrate, here is a sentence from an old legal book:
The amendment consists in the substitution of the word "eighteen" for "twenty-five".
The author is summarizing the amendment—reducing it to its essence. An amendment consists of words: it's made of words. So, saying that the amendment "consists of" a substitution is slightly odd, though it certainly can be done, if you want the reader to think of an amendment not as words but as an action. But in this example, "consists in" avoids the "is composed of" meaning and emphasizes that what follows is the main idea, the important thing to know about the amendment, from which minor details have been omitted. (Also, in older writing, "consist in" and "consist of" overlap much more; there was not such a clear distinction as today.)
When you say "A consists of B", you mean that there is nothing more to A than B; A is to B as a roux is to flour, water, and lard. "A consists in B" is what you resort to when you want to avoid that strong meaning of completeness, but you still want to say that, in a way, even though there is more to A than B, B still captures the whole of A by stating its essence or what is most important about it.
And that's why I can't understand what "Milk consists in vitamin B12" means. Indeed milk doesn't simply consist of vitamin B12; we all know that milk contains many more chemical compounds than just B12. But what special role of vitamin B12 in milk do you have in mind? My mind gives up this search very quickly and concludes that the sentence must be a mistake.
Here's a more typical example of "consists in":
The plot consists in resolving a nest of conflicts arising from Hamlet's father's death.
You could say consists of here, but consists in helps confirm that what follows is meant only as a summary in the most abstract terms. Of course the plot consists of a very specific sequence of events, some of which relate more and some less directly to this structural aspect.
It appears to me, in my admittedly amateur-level understanding, that the English senses of "consist" are remarkably similar to the Latin ones. There are two main differences: in English, the pronouns that compete to disambiguate the common senses are different; and most of the Latin senses have dropped out of contemporary English, such as "I agree", "I'm not changing my mind", "I continue", "I stand still, hold my position, refuse to yield", "I stop and linger", and "I depend on", leaving the sense of "I congeal" echoing most strongly in the various senses of English "consist". English "consist" always takes an indirect object with "of" or "in", though, so "consist" is not a synonym for "congeal", and unlike in Latin, the congealing is already completed and is usually metaphorical. (But see the OED for obsolete English senses of "consist". Most of the Latin senses are represented.)
Dictionary definitions might define "consists of" or "consists in" as "contains", but that just goes to show why you can't trust dictionary definitions. The word "consist" does not evoke the notion of enclosure or containment. You really have to follow the way that metaphors get extended and clarified in real examples of usage. The "congealing" sense is also echoed in, and reinforced by "consistency" when applied to substances, as in "The dough has a firm consistency". It's implied that the dough's firmness is the same everywhere. Even logical consistency draws upon this notion of the same stuff "holding together" throughout: a story is "consistent" if it contains no internal contradictions.
So, the Latin verb consisto carries a mish-mash of meanings centered around conceptually blending hazy notions of standing and holding together. A listener must use common sense and context to understand how the hazy senses are supposed to, er, congeal into a specific sense in any one sentence. English "consist of" and "consist in" are more narrowly focused on the elements that "stand as one" to form a thing or substance distinct from those elements individually.
*Please note that what I'm saying here is not the consensus among linguists.