Uncovering scientific ERRORs: can financial rewards work?

Do you think current ‘ad hoc’ approaches to error detection in the scientific record are sufficient?