No, they won't.
IT departments don't actually have any loyalty to Microsoft. This is important, but ignored by linux zealots who simply don't understand the issues facing IT Professionals. In theory from a certain point of view, there is a window to replace Windows XP with some variety of Linux.
In practice this will not happen. Here's why.
Companies do not run on "Windows", be that XP or whatever.
Companies generally run on two or three applications. First is the word processor, which is not longer an obstacle. Yeah, LibreOffice is perfectly good enough for word processing and it's actually easier to convert users from legacy versions of office to it than it is to Microsoft offerings. (and good job on that, by the way)
Second is an industry specific piece of software, basically CRM or CMS. These are typically tied to word for generating word documents, however it's not so much of an issue as in the past as quite a few are moving towards being web based apps and so therefore platform independent.
Thirdly is Exchange/Outlook, the formidable obstacle that will prevent pretty much any migrations. It's used everywhere, and the integrated mail/calendar is the killer. Before someone says that you can add lighting to thunderbird, I shall point out that outlook allows a delegated user (such as a seceratry) to take a phone call from Mr X, check her principals mailbox to see if he has actioned it and then book an appointment in for that client. This is important, regardless of if that principal is a Doctor, CEO, Solicitor or whatever.
If I was so incautious as to ignore the users wishes on this little issue then I would be seeking employment elseware and my successor would be reimplementing Exchange/Outlook. So, when there is a *working* alternative to Exchange and outlook out there then you can expect to see a huge tidal wave of companies moving away from windows server and Exchange. It's not happened yet, although openchange is getting to within shelling distance of being able to replace an exchange server.
"Working" needs to be defined as gaining user acceptance by being suitable for doing their job, as well as technically operational. If it sort of technically works but the users demand the severed head of the person deploying it then it doesn't work.