I hate those hard lines, they drive me bonkers. That said, it would be possible to open up the ePub and do a find-and-replace feature on
<br />, which should be the equivalent of a new line. It depends on how Calibre's rendered the file in XHTML code. If the paragraphs are also separated by
<br /> however (instead of being styled as
<p> or perhaps
<div>), then doing this causes problems of its own. In any case, it takes a long time and isn't usually worth the effort especially if the PDF's original presentation is poor -- I've tried it once, and never again.
I have an old copy of Acrobat (not Acrobat
Reader), so I'm not sure if it's possible with the latter, but does Reader have a native HTML export option? I've tried it once or twice with Acrobat, and it worked fairly well. No arbitrary breaks, anyway, and formatted properly in regards to paragraphs.
Another option is to crop the margins, and I'm reading a cropped PDF pretty satisfactorily on the Kobo at this moment. Not conversion, obviously, but sometimes it's just not worth the effort. This only works for PDFs with a lot of whitespace (such as academic journal articles -- they often have ridiculously big margins, for some reason -- or the occasional book) and if the font size isn't already on the small side in the PDF. I think you'd need either Acrobat or... I swear there's a freeware program mentioned on MR even, that does the same job.
EDIT: For the removal of margins,
BRISS appears to be one such program. More can be turned up by searching the forums.
Last option I can think of is to use an OCR program that can scan through PDFs and output to an editable format. Not familiar with this route however.
I wouldn't say any of these ideas are necessarily better, but hopefully the brainstorming here will help somebody in some way.