Why is "deducing the architecture of accumulation" on anybody's agenda? This isn't a calculus equation. It's a world where people live under conditions inherited from the past. Central among these conditions is the existence of a money-wielding elite out to perpetuate its established powers and privileges by perpetuating "private" industry.
Within these conditions, the labor theory of value is a category that is useful for keeping track of who puts how much labor into commodities, and who gets the financial and political proceeds attaching to their production. It is useful because it grounds a socialist ethics in the actual, concrete events that are most central to the ethical and practical topic at hand.
Money counts several things, but central, if not primary, among these is numbers of hours of work. A car costs 20,000 times what a hamburger costs because it takes 20,000 times more labor to make a car than it does to make a hamburger. That's not the whole story, but it's certainly not NONE of the story either.
We have a composite theoretical approach to the human psyche and probably all other complex phenomena. Why can't we have one about economic counting? Doesn't reality dictate this? And isn't the labor theory of value one necessary theory there?
Meanwhile, Sweezy and Veblen provide an excellent framework for understanding the political economy of capitalism. This understanding allows us to make some powerful predictions about the future of human events. The relative growth of finance; the importance of military spending; the progressive commercialization of personal life; the main trends in the division of labor; the stonewalling of the welfare state and public industry/employment.
It's fine and dandy to "go beyond" them, but it had better lead to superior explanations and predictions of events. In my book, I see no need for wholesale rethinking. Their model, like all models, needs refinement and elaboration. But it has proven to be extremely robust as an interpretive and predictive framework.