The question is why is the AS-AD model rarely seen on econoblogs.
It seems to me you get an AS curve pretty quickly by assuming that, in the short run, some firms adjust prices and others don't. I mean you don't have to get into Calvo pricing to get to "facing increased demand for a given price a firm might produce and sell more or raise prices so let's assume a bit of both (really some firms don't change prices at all most days but uh aggregate wave hands jump to aggregate demand)".
You have to say something about imperfect competition and second order costs of other than optimal pricing, but really, perfect competition is a strange idea which people don't grasp at first. I think the hard part is the counter-intuition explaining why with perfect competition and flexible prices the AS curve is vertical.
AD has to be given as a function of the money supply. I mean that's the way things are. Yes real world central banks target an interest rate.
I think there is little AS-AD on the web not because bloggers appeal to DSGE, but because most have gone all the way back to an IS curve (real interest and output) assuming AS doesn't matter and with the LM curve replaced with something like a Taylor rule. AS if anything, is an adaptive expectations augmented Phillips curve which matters only because of real interest rates, the monetary authority's response to inflation and debt deflation/inflation.
I don't know of anyone whose policy advice or discussion of current events shows any hint of any intellectual influence of DSGE models. I see basically Friedmanites who say monetary policy is key (I don't understand their arguments) Keynesians who seem to be thinking of an IS curve and a few new classicals who, frankly, seem to have gone quiet lately (by which I really mean neither you, nor DeLong nor Krugman link to them).
I read DSGE in discussion of new Keynesian DSGE models but I can't remember having any sense of anything coming from them beyond aggregate demand matters, expected inflation matters, and there is something like an expectations augmented Phillips curve (except for hysteresis which we will ignore). This is, I think Keynesian economics as of 1960.
OK OK so I ended up dumping on DSGE as usual, but I did think for a minute there.