This has been my experience. I have done this on three sites.
1. Lingerie site originally focused on the keywords "wedding night lingerie" and "Bridal undergarments".
2. Computer repair web site that recommended a certain free computer repair anti-virus, and I switched when the product became a resource hog.
3. A natural pet products site and I had an article on it that I threw together based on inaccurate information about how to control fleas.
In number one, the lingerie site, I dumped all content and revamped the site for Christmass sales, abandoning the Bridal niche, and concentrating on fine lingerie gifts for your lover. Needless to say the site dropped from sight on the bridal words, but the new content was indexed quickly and began to rank fairly well right off. The site still hits for the bridal words, even though they are not on the site anymore, but the rankings are below 200 or so I assume the previous link building is the root of this, as these keywords are not prominent on the site anymore.
Made decent $$ this Christmas from it too. The change was done in the months of SEPT/OCT.
All things considered, the site did not suffer much at all. (I subsequently ruined it by buying a fiverr backlink pyramid gig.)
In number 2, the computer repair web site, I was careful not to change the page content much at all, just made my product changes, and changed the how to instructions to fit the new targeted software. This change was about a 30 percent change in the overall article content, with no real change it keyword density, and meta information. Before the change, the page was PR4, and hitting its keywords. After the change, it was PR4, and still hitting the same keywords. The site showed absouletely no reaction whatsoever. It has been 4 months now, and still looking good.
On site 3, I had to scrap a couple of pages, and redo them with new content. This required a complete change of content and meta information, although the subject matter and title of the page did not change.
In this case, the page took a hit, but the site did not. The new page caused a change in the keyword targeted, so I had to tweak it a little. The new page took a time to start moving up, but it did after a month or to, and settled about where the original page was finally.
Overall, possibly since my sites are usually at least 15 pages, all original content, the sites were able to stand the hit, and they returned to their previous position, keeping in mind that a page you change the content on will usually be targeting a new set of keywords.
If I were you, i would make a correction or two a week until you fix all the wrong information. Change a couple of paragraphs at a time, and allow a week or so between changes. Try to keep your meta info the same if you can.