Pat,
They run the gamut, of course, but I always felt, for reasons Mike Nuzzo describes, that designing a new course, providing its an owner you have synergy with and the "right" amount of involvement is easier than a redo.
As mentioned, in a remodel, existing tree corridors, irrigation, paths and drainage (if you plan to save some of them) begin to proportionally make it less feasible to re-route an existing course. It just starts to get more expensive (with clearing, topsoil stripping, shaping, irrigation, paths, seeding/sodding and drainage. You have to question whether the new hole is really worth the 10-25% extra cost to build over rebuilding in place. It's never a clear black and white decision, but in most cases, the budget factor tends to push you towards keeping most re-routing.
As to topo maps, I use them, of course, but unless there is an unused and heavily wooded portion of the site, in most cases, you can see potential new holes on an existing course without using topo maps. From time to time, I go to existing courses and wonder how the architect didn't see a certain hole. Right now, up in MN, I have re-routed three holes from a 3-4-5 sequence of awkward holes to a much better 5-4-3 routing, each with better views of Lake Superior and each fitting the topo better than before. What's funny is the other holes fit the land so well. Not sure if Don Herfort was off the day those last three were built, or had some constraints (real or imagined) that kept him from what I consider to be a better routing. (We did keep most of the basic corridors)