Building on my work with web2disk, here's web2lrf (part of libprs500
It directly converts websites into LRF files. More than that it has support for profiles that allow it to preprocess websites to generate better looking LRF files. Right now it knows about the New York Times, The BBC, The Economist and Newsweek (see attached demos).
To use it with a profile:
For e.g. for newsweek
For The New York Times
web2lrf --username myusername --password mypassword nytimes
To create your own profile and use it with web2lrf visit
for instructions and examples.
To use it with an arbitrary website (it wont do any preprocessing)
web2lrf --url http://mywebsite.com default