Technically both are correct, but the second version is clumsy- and unnatural-sounding.
Moreover, the structure for the second is not always replicable, whereas the first one is. Consider the following examples:
He sees three people a few hundred meters away.
He sees three of people a few hundred meters away.
Here, using an exact number does not work for the second structure.
Consider another:
He sees several people a few hundred meters away.
He sees several of people a few hundred meters away.
Again, the first structure works and the second one does not.
But the second structure is the only structure that will work when dealing with a word that is only ever a noun. Consider the following two sentences, both of which use the noun "number":
He sees number people a few hundred meters away.
He sees numbers of people a few hundred meters away.
The first is obviously wrong (but note that it can be fixed by converting "number" to "numerous"). And another example with a noun:
He sees group people a few hundred meters away.
He sees groups of people a few hundred meters away.
Both "number" and "group" are definitely nouns, and only the second structure works.
Now consider one last example where instead of using a number (or an undefined number like "several"), we use the word "couple," which can be either a noun or adjective:
He sees a couple people a few hundred meters away.
He sees a couple of people a few hundred meters away.
Notice here that both structures work and sound natural.
So let's return to your initial example. "Thousand" can be a noun or adjective. And like "couple" above, both of the structures you posed work. The second structure is less desirable, but it's technically accurate.
These aren't absolute rules. The English language is treacherous and rarely do the "rules" always apply. But I'd conclude with this: use the first structure, except when using a word that is only ever a noun, when you should use the second structure.
Best Answer
"People came in the thousands" means that thousands of people came. "People came by the thousands" means the same thing, and is equally valid English.
"Crowds swelled by the thousands" means that the size of the crowd increased by thousands of people. The word by is used to compare the size of the crowd before-versus-after the crowd "swelled".
"Crowds swelled in the thousands" is not correct English.