As for your comments about global warming, you're absolutely right. 97% of the world's climate scientists, who are generally not paid very well, agree that global warming is real and a real danger to human existence.
And those 97% of climate scientists would be generally paid nothing if it were determined they'd been flat out wrong (again) for the last 2 decades. At the least they'd be discredited and viewed as incompetent. The grants and foundations that make their work possible would evaporate. A cynical person would point out that pure self-preservation might encourage some to speak that which ensures their job over the truth.
Just saying...
Scientific transparency does not require laying your entire online life open to muckrakers.
It does if the medium used for that ":ife" is funded with tax dollars.
There are many reasons that you're not supposed to use university faculty/government/corporate email systems for personal activity. He's not being asked to provide his personal home email account, nor his personal home computer. The University is being asked to provide access to his tax-funded work email and research data. If he's discussing how he likes to dress up as Little Bo Peep or something in those datasets, then he's misappropriating work assets for personal use, and he's not ensured privacy in that medium. If they really wanted to protect anything but the data itself they could put a gag on those who view the data/emails and disallow them to discuss or share anything not directly related to the court case.
All that being said, I'd love to know how protecting proprietary data or processes in an attempt to "avoid competitive harm" demonstrates the altruism that climate change proponents use as a blugeon against skeptics. If you're so convinced that man-made climate change is a direct threat to humankind and the global ecosystem, is [financial / research / international] competition really a defense for hiding data? Either this is the most dire situation our generation has faced and anything necessary must be done to avoid it, or its not. Apparently, it's not actually dire enough to risk losing research funding to a more "competitive" university.
Variables don't; constants aren't.