> Einstein ... formulated a precise, compact framework that expressed a variety of results that had arrived already in his time
Einstein had an especially productive 1905, publishing three groundbreaking papers among others [1].
His paper on Special Relativity certainly consolidated the work of Poincaré, Lorentz, Minkowski and others, much as you say.
His paper on the photoelectric effect, while building on Planck's work, outright contradicted the earlier work of Maxwell and was a crucial contribution to quantum mechanics.
His paper on Brownian motion, on the other hand, built on very little earlier work and provided the first work offering a method to count molecules; moreover, the paper was the foundation for Perrin's demonstration of the physicality of Dalton's atomic theory.
Each of these three papers had enormous immediate impact, and we still use much of the mathematical forms he introduced in each.
Einstein did not stop with these 1905 papers.
> be summarized in a much more difficult and confusing fashion
Einstein certainly proposed a number of write-downs that were neither easy nor unconfusing. General Relativity (GR) is perhaps a good example of one of those in a theory that has proven to be highly successful. Solving the field equations of GR requires solving a system of ten nonlinear partial differential equations (which in itself would exceptionally difficult even before confronting the explosion into the thousands of elliptical, hyperbolic and undefined terms in the PDEs' couplings in general coordinates). Indeed, we're still confused about how one can decide whether a given solution to the field equations is unphysical and derivative questions like whether a physically plausible solution to the field equations can always be studied using the initial value formulation or something similar.
The mathematical structure of GR is complete and self-consistent, and the notation Einstein developed was extremely concise, but it would be wrong to think that the tersest notation (omitting indices and constant factors) G = T makes GR only as complex as tersest notation (omitting constant factors and in coordinates in which momentum vanishes[2]) E = m of Special Relativity.
---
[1] In 1905 he published more than twenty other papers too, several of which are only "lesser" in that they do not figure in the minds of non-specialists.
[2] People tend to underestimate the complexity of Special Relativity; with suitable choices of coordinates one can use it extremely liberally in flat spacetime, including where acceleration is non-negligible (i.e., where there is obviously non-uniform motion in the systems under study). It's only the presence of real gravity that wrecks the globality of Special Relativity; but even where there is real gravity -- which is never exactly uniform or linear -- Special Relativity is still valid locally).