I assume you are referring just to the postdoc mentoring and scientific communications provisions. I'm not sure how much good the communications provision will do, but I think the mentoring provision is valuable. Here's why:
First, the broad impacts information is used in deciding whether a grant gets funded or not (though I doubt it has very much weight), so people will need to tell a good story. Having a postdoc office at the institution provides an easy out for PIs - they can just write something about the things the office will do for the postdoc. If offices are smart, they'll distribute boilerplate text that can be used for just such a purpose. That in turn I think will generate support (or at least grudging tolerance) within institutions for postdoc offices, which are by no means universal. I think it will speed creation of such offices.
Second, it makes an explicit statement that mentoring is an expected part of the postdoc experience. That's a small step towards formalization and professionalization of the postdoc. Ultimately, I think the postdoc will evolve toward a grad-school-like structure. There will be different levels of structure, of course, and different milestones and expected tasks, but still clear goals and steps towards those goals. I think it's a good direction to be heading in.
I assume you are referring just to the postdoc mentoring and scientific communications provisions. I'm not sure how much good the communications provision will do, but I think the mentoring provision is valuable. Here's why:
First, the broad impacts information is used in deciding whether a grant gets funded or not (though I doubt it has very much weight), so people will need to tell a good story. Having a postdoc office at the institution provides an easy out for PIs - they can just write something about the things the office will do for the postdoc. If offices are smart, they'll distribute boilerplate text that can be used for just such a purpose. That in turn I think will generate support (or at least grudging tolerance) within institutions for postdoc offices, which are by no means universal. I think it will speed creation of such offices.
Second, it makes an explicit statement that mentoring is an expected part of the postdoc experience. That's a small step towards formalization and professionalization of the postdoc. Ultimately, I think the postdoc will evolve toward a grad-school-like structure. There will be different levels of structure, of course, and different milestones and expected tasks, but still clear goals and steps towards those goals. I think it's a good direction to be heading in.