Re: @ Korev
Munich use Kolab as a back end. I'm not sure how it compares point to point with Exchange but the relevant issue would be how well it suits their workflow. AIUI they worked pretty closely with the Kolab developers and for all I know still do so that's not likely to be an issue.
The issue isn't the back end. Kolab supports both the Open Standards imap / smtp / caldav / carddav / webdav series as well as access through ActiveSync licensing, which Outlook uses, and most other Groupware platforms like OpenXchange and Zimbra do the same. The problem is the front end. Astonishingly, there is nothing that offers the functionality that Outlook offers, at least not something that doesn't look like it's last design change was in the 90s.
Like it or lump it, there are design trends too and that matters to end users, but the bit I find interesting is that a MICROSOFT piece of software has not yet been replaced with something better in FOSS. I've tried Thunderbird (hence the 90s comment :) ) but it's a fight to make it do decent HTML/RTF emails which is what users want (no, not need, want).
I've added Lightning, but unlike other calendar apps I can't subscribe to a caldav server in a way that it auto-adapts like, for instance macOS can - if I create a new calendar macOS will pick it up, TB Lightning needs me to feed a new full URL. And unless you install an extra plugin (Lightning Calendar Tabs) it's not world's best functionality either.
I've tried both the CardBook and the Inverse SOGo connector plugins to add a carddav feed, and decided on the SOGo one because its UI integrates better (but here too is a UI question: why can't the address book just be a tab too?). To be honest the carddav aspect was the leats painful, that just pretty much worked out of the box.
Oh, and then you have the next issue: if you get TB to work it will not talk the proprietary protocol for Exchange, and you need both client and MTA sides before you can reasonably expect to start making a difference.
In short, there's a lot of heavy lifting required to replace Outlook with something that is as usable, and the blunt answer is that nothing exists in the FOSS world - not for Linux, not for macOS and not for Windows. As a matter of fact, apart from *cough* Evolution *cough* (which is IMHO terrible) nobody has even made a *start* looking at this, it's clearly not sexy enough to attract funding or talent, so that's game over.
Biting the hand that feeds IT
