Traditional Keynesian models, like the IS-LM one, were discarded as they lack micro-foundations. They were also subject to the Lucas critique (in the sense that one cannot properly estimate a macroeconomic model if the parameters respond to changes in monetary/fiscal policy).
Quite surprisingly, although there is so much talk about the liquidity trap and its close concept, the zero lower bound (see the definition of liquidity trap), the criticism of these concepts is rather thin. This is even more puzzling since the liquidity trap concept is known for a long time, ever since Keynes proposed it (Rhodes did not find any mention of it in the work By Keynes).
Explaining the crises (not all, but many of them) as being liquidity traps is not only a misinterpretation but it also leads to false solutions. Just look at the case of Japan after two decades of “policy experiments”. (New) Keynesians like Krugman have reduced its stagnation problem to a liquidity trap and prescribed a wrong therapy which in the end failed to lead to real economic growth. But probably the case of Japan deserves a separate discussion.
There is a general agreement that the mainstream macroeconomics has largely failed to predict or correctly estimate the last economic and financial crisis, see the introduction here on the website of the Institute for New Economic Thinking. This failure has motivated more than ever different heterodox approaches to macroeconomics.