Personally, I find monospaced, serifed fonts to be the easiest to read, and am frequently frustrated by the lack of books that use a monospaced font, and I wonder who is to blame for it.
In all of the software usability testing that I've done over the years (mostly involving web applications), monospaced fonts consistently score the highest for readability and comfort with the users.
(Also, tables etc in plain text need a monospaced font to line up properly.) I would bet that the reason publishers like condensed text is the same reason they indicate new paragraphs by indenting the first line rather than separating adjacent pars with a blank line; the same reason they will break a long line of metrical drama and append the tail to the end of the line above or below; the same reason they will publish a long narrow list as a two column table even though there's no connection between the left and right halves of the content on any one row: paper costs money, and they want to get as much use out of each square inch as possible while keeping the text more-or-less readable. They are largely happy to sacrifice some readability if it means they can limit the amount of paper they have to use. Of course, in etexts we don't have that problem. Screen space (especially vertical screen space) is as good as free. Using a blank line makes it easier to see where a new paragraph starts? Fine, we'll do that. Rejoining metrically split lines? We'll do that too. And since in plain text files it's trivial to choose, I always read PG texts in a monospaced font. Preserving the financially-induced presentation limitations of publishers and printers working with paper in etexts seems a waste of effort and introduces an unecessary loss of readability, in my opinion. People who care about the exact typography or detailed layout etc will one day be able to go look at the scans - people who just want to read/search/edit/quote the actual textual content don't need all the bells and whistles, right? :) Cheers! Bill