Nothing that leads to logical inconsistency is "confirmed by evidence". Expansion leads to logical inconsistency analytically. To wit, if there were something outside reality that were sufficiently real to contain the "expansion" of reality, it would be contained in reality. That's a contradiction; ergo, the hypothesis is false.
The overall size of the universe is externally undefined and can only be defined intrinsically (as curvature), the sizes of objects change with respect to this curvature.
The cosmos can’t be expanding in any absolute sense, because there’s nothing for it to be expanding into. Therefore, we must invert the model in a way that “conserves spacetime”; the total “amount” of spacetime must remain constant. When we do so, the cosmos ceases to resemble a balloon inflating (extending outward) over time, and instead becomes an inward superposition of sequentially related states. The best way to think of it is in terms of a cumulative embedment of Venn diagrams (of state) on the inside surface of a sphere of extrinsically indeterminate size.
"Intrinsic expansion" is a contradiction in terms. If something is expanding, then it has to be expanding *with respect to* a fixed referent, and if it is, then it has to be extending into an external medium with respect to which the fixity of the referent has been established. On the other hand, saying that something is shrinking relative to that which contains it presents no such problem, for in that case, nothing is really "expanding". An inclusive relationship, like that whereby the universe includes its contents, can change intrinsically only if its total extent does not change; where its total extent is just that of the inclusive entity, this means that the extent of the *inclusive entity* cannot change. Ergo, no expansion; it's logically analytic. Reason in any other fashion, and the term "expansion" becomes meaningless.