His statement was that since because people cannot see the infinitely small changes in their preferences of utility the calculus we use to map a continuous function of utility is both useless and false. My thoughts on the matter are that Bryan is correct. When it comes to the supply curve meeting the demand curve it will only happen by chance and in any real situation the model will fall apart under close inspection. I do believe though that regardless of it not being the end all be all truth in economics there is merit to the idea of a utility curve and that simply because we can't graph a humans desires doesn't mean it isn't helpful in estimating/guessing the utility of said item.
Overall the arguments Bryan brought forth were good but took the work of Austrian economics to their logical extreme which I feel is pointless.