I'm looking for a reference for analogues of the Blakers-Massey triad connectivity theorem (and its higher-order generalization) for ring spectra. That is:
Suppose that $A\to A_1$ is a $k_1$-connected map of (associative) ring spectra and $A\to A_2$ is a $k_2$-connected map of ring spectra, and that the maps are cofibrations so that the pushout, call it $A_{12}$, is a homotopy pushout. Then as long as all of the rings are connective the map of spectra $$ A\to holim (A_1\to A_{12}\leftarrow A_2) $$ is $(k_1+k_2)$-connected.
Has anyone worked out a detailed proof of this? I would be happy to see this in any reasonable theory of structured ring spectra.
EDIT: I know how a proof should go: Filter the spectrum $A_{12}$ by "word length" and examine the sequence of subquotients (the "associated graded object"), which are wedges of "tensor products" of the spectra $A_i/A$ regarded as bimodules over $A$. But I do not want to delve into technicalities if the details are already out there somewhere.