#2... I dont' know, that's what Wall & Jordan's book says. Basically just don't expose the film during the clearingtime. But it seems like a useless suggestion, right?
#1... You could be entirely right. I guess a could place to start would be between 2 & 5 g/L, encompassing both author's suggestions.
I looked into D-19 a bit (new to me), and it's a high-contrast, high-grain, scientific type developer. It has high capacity and relatively good shelf-life. It makes sense that this would be a 1st developer, since it's creating a pretty hard negative image, thus the positive should be nice and punchy, with low grain.
I can take another look this evening and try to clarify some of these recommendations.
edit: The impression I get about the silver solvent is not that its only purpose is speed of processing, however that is a concern for movie film as noted by Haist. The real reason is for physical development. Here, check out this (there was a url link here which no longer exists) Not to mention, the solvent acts differentially, and complentary within the image, that is, it works where it needs to and not where it doesn't. I'm afraid that a farmer's-type reducer would decrease overall density, as opposed to preferentially in the highlights. That seems like a lot more fuss to me. Additionally, it isn't mentioned in either text, though that's not to de-validate it by any means. I am just curious where you learned or discovered it.