The paper sheds useful light on an issue that is faced by operational forecasting services related to trade-offs between forecast quality and timeliness. On the whole, I support the revisions suggested by the reviewers, which will require a minor effort to address, and I find the authors' responses to their comments to be generally appropriate. I suggest that the authors proceed with the corrections that they have outlined, although I would encourage the authors to consider the additional figure suggested by one reviewer, in which the 0-day forecast is used as the reference to show the distribution of skill changes across the catchments and seasons -- even though the following figures summarize skill change distributions.