Grant and fellowship applications are the most difficult thing to write. It is easier to do the actual scholarship, in many cases. I spent a few hours last night helping out someone with one of those, as a favor (though I will also do it for a stranger for a fee). Here's how I approach this kind of task.
The most difficult thing is to address a broad audience. In our scholarship we address other experts, but for a fellowship we need to emphasize the significance of the project, why anyone outside the narrow sub-speciality should care. Part of my task, then, is to help the person clarify specialized terms. Some people make a ritual gesture in saying that their work will be of interest to those in Comparative Literature, Sociology, disability studies.... In other words, they make a list of fields that are tangentially related to their project. This does not work unless there is a specific link articulated to one of these fields. For example, just because your work deals with performance does not mean that those in performance studies will be interested. You have to say what contribution you are making to this other sub-discipline. For example, you could make the argument that disability studies has never considered the influence of a certain cultural dynamic that your work brings out.
I've often said that people think of themselves as more anti-canonical than they really are. To write about obscure writers requires extra justification. And by obscure, I mean almost any writer at all aside from the few supercanonical ones. Nothing is supercanonical in my field except maybe Borges, Cervantes, and Lorca, and maybe a few boom novelists who've won the Nobel prize.
Since proposals are very condensed, it is difficult to avoid some vagueness and lack of clarity. At the same time, I am very harsh with vagueness in my comments, since this is a potentially fatal flaw. You can be concise and still precise. A summary of chapters of the proposed book is always helpful, with as much detail as possible.
Theoretical language, or jargon, is usually not welcome in grant applications. I usually tell people to cut that to a minimum.
It is good to read a proposal from the point of view of ignorance. In other words, the less I know, the more I can put myself in the position of the average historian or English professor with no specific knowledge of the topic. Of course, I won't necessarily catch mistakes that would be visible to a specialist reader, but a specialist might not realize what is unclear to the non-specialist.