I thought this was for on premise exchange installs that are directly facing the internet, which is an extremely rare setup these days.
Most companies use hosted exchange or if exchange is on premise, it sits behind a firewall of some kind.
Exchange is actually still fairly prevalent, even among smaller companies. Although many of the smaller orgs that still have on-prem Exchange tend to have a migration plan to M365.
and I hope they do. most of these smaller companies are sometimes sitting on really really old versions. "it works" is mostly the argument. updating exchange sometimes can be painful. most of the time everything works, but sometimes things just break.
There are vanishingly few circumstances where it makes sense for an organization to be funding deep expertise for the direct management of an Exchange environment. This has been clear for nearly a decade.
The capex to refresh that hardware is a ridiculous waste, so yeah, it wouldn’t surprise me if the people still running those setups have very aged installations (e.g. WinSrvr 2008-12), which are as great a risk as the Exchange Server software they’re running.
The gating factor is often the expertise to plan and execute a migration with minimal disruption and loss. It’s not simple, and it’s nothing like an exchange upgrade project. It’s a downright UGLY project if a company has been abusing their mail system for years (e.g. using their mail system as a document management platform since ‘99, allowing distributed PSTs, etc.). Seen it.
The only signal I would conclude from CVE data by itself, is that I bias towards a preference for companies that regularly publish CVEs. The ones that don't publish CVEs regularly are hiding, ignorant, or actually secure (and the first two are more likely).
You can't look at CVE in isolation.
Possibly if a product consistently has high cves over a long period of time that might tell you something about poor security practices over that period (or before it). It might also mean that their security is now quite good!
You have to interpret the data I'm afraid. I can't think of any useful statistical measures you could use to compare aggregate data across multiple products.
Give Microsoft money and receive Teams for free?
Yes. As opposed to giving Microsoft the same money and then also giving Slack some money too.