Basically the RBC model is based on stochastic growth model. Which of course is extended by including the household choice of leisure and work.
Prescott and Kydland (1982) proposed a model including time-to-build feature and unseperable leisure, this model fits the second moment of actual data very well except some labor market varialbes. Hansen(1985) introduced a indivisible labor and labor horading model to explain the high volatility of working hour. The later development of RBC includes addding more shocks, like government spending, or introduces distortionary taxex.
Marjor critic to RBC model includes: Cogley & Nason(?) point out that the RBC model is lack of propogation mechanism, all the autocorrelation and trend reverting feature comes almost wholy from the techonology shock, rather than endogenously generated by the model. Gali (1999) points out that the RBC model made wrong prediction of how techonoly shock affect the labor hour, labor productivity etc, he then preoposed a sticky price model and a SVAR, which empirically rejects the RBC prediction of the labor market dynamics.
And of course, the new Keynasian model, which emphasizes more on monetary shocks disagree with RBC model in the source of the economic fluctuation.