"For example, even with Cortana and searching the Web from the Start menu disabled, opening Start and typing will send a request to www.bing.com to request a file called threshold.appcache which appears to contain some Cortana information, even though Cortana is disabled. The request for this file appears to contain a random machine ID that persists across reboots."
and
"if Web searching and Cortana are disabled, we suspect that the inference that most people would make is that searching the Start menu wouldn't hit the Internet at all. But it does. The traffic could be innocuous, but the inclusion of a machine ID gives it a suspicious appearance."
That's as personally identifiable as you need when you have an identifier locked to the machine that is sent even across reboots. This is even more identifiable than an IP address, even for static IPs (since you can have multiple machines behind it).
Is that sufficient for the claim or do you still reject it? I'm most concerned that you hold anger at FOSS people's "arrogance" over complaints about a bad software package that many people are unfortunately hooked to.
Anti-FOSS advocates here in these comments are behaving poorly; a conclusion they don't like which is supported by facts that are referenced isn't being taken well.
Here's some examples of where it is going wrong in these comments: attacking the messenger, claiming things are false / unsupported when they are, trying to find any minor flaw in what was stated in order to discredit the whole thing, ignoring facts and misleading people about what the source actually says.
That's some nasty stuff. That's Us vs. Them stuff, not "yeah that is a concern, I'm upset about that too!"