It feels like this proposal maintain the paper as the central artefact, but maybe the paper should be the side/byproduct. In many computer science fields it is clear that reproducible code should come first for example.