I disagree on your first point- I think that there could be someone else who could write a grant proposal for me, provided that they a) expressed an interest in

I disagree on your first point- I think that there could be someone else who could write a grant proposal for me, provided that they a) expressed an interest in grant writing over benchwork / data analysis and b) were fluent enough in the field's literature to be able to translate my ideas into bureaucratic prose more elegantly and efficiently than me. Such a hypothetical individual might have some overlap with the "scientific popularizer" roles that we see cropping up with figures like Neil deGrasse Tyson or Bill Nye, in the sense that they could serve as a conduit between scientists and funding (whereas popularizers serve as links between scientists and the general public). The issue is that this kind of idealized division of labor can't exist in the current climate- there just isn't enough money to distribute a researcher's load across multiple specialists. This results in researchers being expected to juggle advising, grant writing, committee work, publishing, coding, analysis, and teaching, which creates more stress on researchers and poorer quality results. I do agree with you on your second point though- publish or perish is academia's equivalent to rank and yank.