Between model theory and category theory broadly conceived: not anything really compelling, because a category, on its own, does not stand as an interpretation for anything.
Between model theory and categorical logic, however: yes, I think the overlap is large.
A spot of history: the man most deserving, in my opinion, of being called the father of model theory is Alfred Tarski, who came from a Polish school of logic that, I understand, was very much within the algebraic school. His model theory was more in the vein of a reworking of the Polish-style algebraic logic (this is not, in anyway, to talk down his achievement).
Blackburn et al (2001, pp 40-41) talk of a might-have-been for the Jónsson-Tarski representation theorem:
...while modal algebras were useful tools, they seemed of little help in guiding logical intuitions. The [theorem] should have swept this apparent shortcoming away for good, for in essence they showed how to represent modal algebras as the structures we now call models! In fact, they did a lot more that this. Their representation technique is essentially a model building technique, hence their work gave the technical tools needed to prove the completeness result that dominated [work on modal logic before Kripke].
They go on to present a nice anecdote showing how Tarski did not seem to think this algebraic approach provided a semantics for modal logic, even after Kripke stressed how important it was to Kripke semantics. It seems that sometimes algebraic logic and model theory are more similar than they appear.
Like model theory, categorical logic can seem to be a special way of doing algebraic logic. And with some theories, model theory and algebraic logic sometimes seem to differ only in trivialities; with categorical logic I am more hesitant in making sweeping judgements, but it sometimes feels that way to me too.
Ref: Blackburn, de Rijke, & Venema (2001) Modal Logic, CUP.
The categories of models with elementary embeddings are accessible categories. (The cardinal κ is related to the size of the language via Löwenheim-Skolem; the κ-presentable, aka κ-compact, objects are models of size less than κ.) Michael Makkai and Bob Paré originally describe this idea in Accessible categories: the foundations of categorial model theory (Contemporary Mathematics 104, AMS, 1989). However, still more can be found in later works such as Adámek and Rosický, Locally presentable and accessible categories (LMS Lecture Notes 189, CUP, 1994).
More generally, abstract elementary classes can also be viewed as accessible categories. Thus accessible categories include categories of models of infinitary theories, theories with generalized quantifiers, etc. In fact, accessible categories can always be attached to such structures, but I don't know the exact characterization of the categories that arise from models of theories of first-order logic. The Yoneda embedding can sometimes be used to attach first-order models to accessible categories, such as when the accessible category is strongly categorical (Rosický, Accessible categories, saturation and categoricity, JSL 62, 1997). On the other hand, you can reformulate a lot of model theoretic concepts in general accessible categories. There are more than a few kinks along the way and not all of it has been done, but the more I learn the more I find that this is actually a very interesting and powerful way to approach model theory.
Let me try to explain the situation in greater detail. I guess the correspondences are better explained in terms of sketches. (This nLab page needs expansion; Adámek and Rosický give a nice account of sketches; another account can be found in Barr and Wells.) A sketch asserts the existence of certain limits and colimits, or just limits in the case of a limit sketch, taken together these assertions can be formulated as a sentence in L∞,∞ (sketchy details below). Like such sentences, every sketch S has a category Mod(S) of models. Sketches and accessible categories go hand in hand.
If S is a sketch, then Mod(S) is an accessible category, and every accessible category is equivalent to the category of models of a sketch.
If S is a limit sketch, then Mod(S) is a locally presentable category, and very locally presentable category is equivalent to the category of models of a limit sketch.
When translated into L∞,∞, a limit sketch becomes a theory with axioms of the form
$\forall\bar{x}(\phi(\bar{x})\to\exists!\bar{y}\psi(\bar{x},\bar{y})),$
where $\phi$ and $\psi$ are conjunction of atomic formulas (and the variable lists $\bar{x}$ and $\bar{y}$ can be infinite). When the category is locally finitely presentable, then these axioms can be stated in Lω,ω. Theories with axioms of this type are essentially characterized by the fact that Mod(T) has finite limits.
If T is a theory in Lω,ω and Mod(T) which is closed under finite limits (computed in Mod(∅)), then Mod(T) is locally finitely presentable category (and hence finitely admissible).
Every locally finitely presentable category is equivalent to a category Mod(T) where T is a limit theory in Lω,ω (i.e. with axioms as described above).
It is natural to conjecture that this equivalence continues when ω is replaced by ∞. Adámek and Rosický have shown in A remark on accessible and axiomatizable theories (Comment. Math. Univ. Carolin. 37, 1996) is that for a complete category being equivalent equivalent to a (complete) category of models of a sentence in L∞,∞ and being accessible are equivalent provided that Vopenka's Principle holds. In fact, this equivalence is itself equivalent to Vopenka's Principle. (It is apparently unknown whether accessible can be strengthened to locally presentable.)
Now, if T is a sentence in L∞,∞, then the category Elem(T) (models of T under elementary embeddings) is always an accessible category. The category Mod(T) is unfortunately not necessarily accessible. When translated into L∞,∞ sketches become sentences of a special form. A formula in L∞,∞ is positive existential if it has the form
$\bigvee_{i \in I} \exists\bar{y}_i \phi_i(\bar{x},\bar{y}_i)$
where each $\phi_i$ is a conjunction of atomic formulas. A basic sentence in L∞,∞ is conjunction of sentences of the form
$\forall\bar{x}(\phi(\bar{x})\to\psi(\bar{x}))$
where $\phi$ and $\psi$ are positive existential formulas.
- A category is accessible if and only if it is equivalent to a category Mod(T) where T is a basic sentence in L∞,∞.
It would be great if one could simply replace accessible by finitely accessible and sentence in L∞,∞ by theory in Lω,ω, as in the locally presentable case above. Unfortunately, this is simply not true. The category of models of the basic sentence $\forall x\exists y(x \mathrel{E} y)$ in the language of graphs is accessible but not finitely accessible. A counterexample in the other direction is the category of models of $\bigvee_{n<\omega} f^{n+1}(a) = f^n(a)$, which is finitely accessible but not axiomatizable in Lω,ω.
Best Answer
Here is a negative answer to question 2 and the converse of question 3.
Let $T_1$ be the theory of the integers under successor $\langle\mathbb{Z},S\rangle$. This theory asserts that $S$ is bijective and has no cycles of any finite length. That theory is complete, since all models of size $\aleph_1$ are isomorphic, consisting of $\aleph_1$ many $\mathbb{Z}$-chains, and more generally there is only one model of uncountable size $\kappa$ for any uncountable $\kappa$. But there are countably infinite many countable models up to isomorphism, since every model consists of some countable number of $\mathbb{Z}$-chains. So $I_{T_1}$ is the function that says there are $\aleph_0$ many countably infinite models, but only one model each in any uncountable cardinality.
Let $T_2$ be the theory of infinitely many distinct constants $c_n$. There are countably infinitely many countable models of this theory, depending on the number of objects in the model that are not the interpretation of any $c_n$, but there is only one model up to isomorphism in any uncountable cardinality, since any model of size $\kappa$ has the distinct $c_n$'s and then $\kappa$ many additional points.
Thus, both of these theories have the same number of models in any cardinality: no finite models, $\aleph_0$ many countable models, one model in each uncountable cardinality. So $I_{T_1}=I_{T_2}$.
But $T_1$ is not interpretable in $T_2$, since if it were, the interpretation would involve only finitely many constants, but one can define only finitely many points in a model of the reduct of $T_2$ to those finitely many constants plus one parameter, whereas in any model of $T_1$, we can define infinitely many points relative to any parameter. So this is a counterexample to the converse of question 3.
Note also that $T_1$ has no rigid models, since every $\mathbb{Z}$-chain admits translations, but $T_2$ has two rigid models, namely, the model with only the constants, and the model with the constants plus one more point. (Also, $T_2$ has models with only finitely many automorphisms.) These features show that $\text{Mod}(T_1)$ and $\text{Mod}(T_2)$ are not equivalent as categories, so this provides a negative instance to question 2.