Before gravity was conceived as a physical phenomenon and studied by Isaac Newton, gravity was not microfounded. So before Newton, modelling the real world should assume away gravity, because it was not microfounded. We should instead create models into which humans could walk on air.
When a phenomenon appears stable, while attempting to understand and decipher it (to micro-found it), we should be wise enough to accept its existence and include it in our models, of course in an "ad hoc" manner since we haven't decipher it yet.
Whether firms "cannot" or "choose optimally not to" change their prices everyday, makes little difference, as long as we haven't settled on the optimization framework into which such an optimal choice could come about. It makes little difference because in the absence of knowledge it can be seen as a semantic issue: I can re-interpret "cannot" as "an optimizing agent cannot do what is suboptimal -it goes against every inch of its existence".
Nevertheless, bashing for micro-foundations of an observed regularity (and it is a prevailing observed regularity across time, space, cultures and economic systems that firms don't changes their prices continually), is a good thing, in that it keeps up the pressure to finally find those microfoundations.
I just hope the bashers don't argue in addition, that in the meantime, we should only be modelling people as walking also on air -sorry, be modelling firms only as being able to change their prices continually.
And to send the ball back, I have never seen in my life any convincing "micro-foundation argument" for the assumption that prices are perfectly elastic.