Posting to RDA-L
Concerning abbreviations, there are an entire range of options today instead of the rather atavistic method of retyping everything. I personally think automated methods, plus using our MARC fields and language of the item would solve at least 90% of all of the "abbreviation problem". Many abbreviations are only valid in certain fields, e.g. see Yale's list of (uh-oh!) AACR2 abbreviations for a nice overview: http://www.library.yale.edu/cataloging/abbrev.htm.
Other ideas come from sites such as http://www.abbreviations.com/, which has different methods for finding out the meaning of an abbreviation from widgets to iPhone apps. They also have an API that can work as a web service. If the library world did something like this, it could solve the "abbreviation problem" not only for English-speaking people, but for everyone everywhere, no matter what language they speak.
This is, of course, assuming that there actually is an "abbreviations problem" and that it is of sufficient import that we must take major efforts to solve it. Whether this is true or not is another matter, but it only makes sense to at least try some automated methods before embarking on a major task of manual retyping.