Wow, 9261, you are right! And it is even crazier that it is random which ones get to change their prices, totally insane. Clearly the New Classical RBC models are much superior with their modeling of market power of firms to figure out which ones are more likely to have sticly prices in contrast with the ones in the more competitive markets that do not have sticky prices.
NC RBC models are definitely way more realistic with their well-specified exogenous shocks and their assumptions of rational expectations. Every piece of evidence we have from surveyst to experiments to just plain old common sense tells us that everybody has rational expectations!