While data isn't copyrightable, the published article is. Scientific journals are not just collections of spreadsheets and data files, they are a curated, peer-reviewed, filtered, and edited product intended to deliver a cohesive set of information to specific audiences. That takes effort and resources, which means that resulting product is a copyrightable work of art. If you stumble across a data file hosted on Joe Smith's faculty website while searching for info on your topic, how do you know if it's any good? If that same information appears as a peer-reviewed article in a prestigious journal, it could have a different value for you -- it's been vetted. That process doesn't happen magically without resources. If the resulting product is made freely available, the resources come from elsewhere (eg, submission fees), but it's still copyrightable.

You're assuming that all "companies" are for-profit and are making money. What about non-profit organizations or those companies that are underwater financially?

2) Microsoft quit attempting to make all of their file formats dependent upon the OS/software that they write. The data is MINE, and I should be able to use other software to read the data. Commit to open file formats and I'd look a lot more favorably on MS.

That's the real trick, isn't it? To devise a file format that is completely independent of any editing applications and whatever features/limitations they happen to have?

Folks have been trying to do essentially that on the web for a decade, and there still isn't a uniform, file-renders-exactly-the-same-on-all-platforms format.

It isn't a Microsoft issue, it's an industry-wide issue.