Justin, I am not an anti class theorist. You know better than that. In fact, class features prominently in my academic work.
All I am objecting to is one-size-fits-all explanations. Any concept, class including, can have a good explanatory power in one set of circumstances, but not necessarily in others. For example, trying to explain European industrialization or welfare policy without taking into account class is crap science, purse and simple. But class does not have such a good explanatory power in the US in this respect, and for good reason, as Theda Skocpol (_Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States_) convincingly argued. In a nutshell, her argument is that machine politics and gender politics in the US cut across class lines, rendering class-based actions (a powerful factor in Europe) ineffective. I would also add race, ethnicity and religion to this mix.
Another point is what is it exactly that we want to explain with the concept of class? Wages and occupational mobility? Or political preferences and voting patterns? It is true that class explains a lot - as you say (albeit I would rather have it expressed as an R square) - but that is a bit tautological, no? After all, class is defined by occupational status and wages.
But explanatory power of class in US politics is not that great, to say the least. People who work for a living (i.e. the working class) - have a wide gamut of political opinions, voting patterns, socio-cultural preferences and so on. Even the wealthy elites are split in their political preferences. Even the whole concept of "working class" is awfully murky and tainted with cultural identities - a truck driving and deer hunting autoworker is "working class" whereas a lexus driving and coffee sipping computer technician is a "yuppie." This is not social science in any true sense, but identity politics of the worst kind.
Wojtek