Two reasons. (1) Physical simplicity doesn’t equate to conceptual simplicity. (2) Whether a given statement is to be regarded as complex, and how exactly it is to be regarded as complex, depends on the dialectical context.
Explanation of (1): Suppose for argument’s sake that there are minimal units of physical composition, these being maximally simple building blocks of nature. Given only that these entities are physically simple (or at least relatively so), it doesn’t follow that the truths that describe them, and in terms of which we are to understand them, are correspondingly simple. If anything, the opposite holds: because these entities are known in a such a theory-mediated way, those truths are exceptionally complex. So even if there are ‘atoms’ (minimal building blocks) of nature, they are not building blocks in a logical sense—on the contrary, they are, in that respect, at the opposite extreme of the epistemological pyramid.
Explanation of (2): Consider the statement “1+1=2.” That statement could function as an axiom (an accepted truth that neither requires nor admits of justification) or (as is the case in set theory), it could function as the end-result of a long chain of reasoning. Any given truth is entailed by infinitely many others and itself entails infinitely many truths. And since a statement’s entailment-relations (what it entails and what it entails it) reflect its structure, there are no ‘ultimate’, foundational truths. Consequently, there are, in this sense, no ‘atoms’ of nature—-no non-composite truths out of which other truths can be constructed or from which they can be derived.
Summary: Atoms, in the sense in which the philosophy of logical atomism requires them to exist, don’t.