To respond to only a portion of James Babb's series of good questions, I think
that a key to LDP strength had more to do with residential stability than
rurality per se. At the individual level, Flanagan and Richardson found long ago
that the correlation with length of time living in one place with LDP voting was
very high. At the community level, it is relatively stable places where one
would see the normal-type chonaikai, largely based on local shops and nearly
always attached to the LDP, with the deepest roots and the most influence on how
the locals vote. The persistence of LDP strength in "rural" areas that actually
had been quite tranformed in their economic and social bases rests on this
point, I think. Both the long-term decline of the LDP vote and the levelling off
from the mid-1970s match the urbanization trends very nicely (as Babb notes,
urbanization declined sharply in the early 70s sometime).

As for what happened to the LDP then, Flanagan's last chapter in _The Japanese
Voter_ has the story through the 80s, on the vulnerability of the urban part of
the party's vote. In part it is the fragility of support based on valence
issues, but that is an entirely different story.