Don't misrepresent us--we support and publish many open formats, but, yes, we do want to have a plain text version of everything possible.
We're looking at our history, and we're planning for the long term--the very long term.
Today, Plain Vanilla ASCII can be read, written, copied and printed by just about every simple text editor on every computer in the world. This has been so for over thirty years, and is likely to be so for the foreseeable future. We've seen formats and extended character sets come and go; plain text stays with us. We can still read Shakespeare's First Folios, the original Gutenberg Bible, the Domesday Book, and even the Dead Sea Scrolls and the Rosetta Stone (though we may have trouble with the language!), but we can't read many files made in various formats on computer media just 20 years ago.
We're trying to build an archive that will last not only decades, but centuries.
The point of putting works in the PG archive is that they are copied to many, many public sites and individual computers all over the world. No single disaster can destroy them; no single government can suppress them. Long after we're all dead and gone, when the very concept of an ISP is as quaint as gas streetlamps, when HTML reads like Middle English, those texts will still be safe, copied, and available to our descendants.
The PG archive is so valuable, yet free and easily portable, that even if every current PG volunteer vanished overnight, people around the world would copy and preserve it.
If the ZIP format loses popularity, and is replaced by better compression, it will be easy to convert the zip formats automatically (and we post all plain-text files in unzipped format as well). If hard drives are replaced by optical memory, it will be easy to copy the files onto that. If even ASCII is superseded by Unicode or one of its descendants, it will be possible for our grandchildren to convert it automatically (and ASCII is included in Unicode anyway).
By contrast, many of us have files saved in proprietary formats from word-processors only 5 or 10 years old that are already impractical for us to read. Some of our files produced just a few years ago using non-ASCII character sets like Codepage 850 are already giving problems for some readers. Some eBook reader formats launched within the last few years are already obsolete. We have learned from that experience.
We also encourage other open formats based on plain text, like HTML and XML, and even occasionally not-so-open ones when simple formatting isn't enough, but plain text is the only format we're sure of in a rapidly-changing technological landscape.
Please see also the FAQ [F.1] "What formats does Project Gutenberg publish?" for more detailed discussion of formats.