I got the impression that indeed mathematics is done exactly this way. People used to write proofs, somebody would find issues in them that don't invalidate the entire work (or they do), the mathematician would work to correct the issues and resubmit.
