Thanks, that's clearer. Let me clarify also on my side. I did not think that any of the tasks that @Martin originally mentioned implied that the grant would help ship improvements. My concern was precisely that the grant application was unambiguously not focused on software development. I think we need to create more capacity (be it via funding or other means) for software development in this area. So, making the grant application be more distant from software development does not make it more appealing to me.
Currently we have @Sebastian working on improvements to the Wikibase integration, with a narrow focus on Wiki Commons. That's fantastic, but the grant that funds this work is rather small and will expire soon. I have also made some improvements in that area recently (with the error reporting changes, a byproduct from the reproducibility project), but I'm sun-setting my activity in the project as a whole. That only leaves @tfmorris, who considers himself unfamiliar with the Wikibase integration and reluctantly reviews PRs on this part of the project. So, as things stand, in 6 months we may well not have anyone actively working on this integration as a developer.
So, assuming this grant gets awarded, my worry is that will have a great community manager talking to the Wikimedia community, attending conferences, hearing all the needs and wishes people have for this integration, but no-one to actually make changes to the tool. I am worried that this is going to be a frustrating job for this new hire, and that the community will get impatient. Given the public feedback we got at the OpenRefine meetup at Wikimania from Asaf (that funding to improve the OpenRefine-Wikimedia integration can be found), it's likely that the community will find it difficult to understand why the project does not just do that.