The sources are always relevant - I do not really believe in "unbiased" sources - I recommend always checking news with multiple different sources who got different interest in a case. In this case the sources are ZDNet and Microsoft.
In this case the story give a different picture to what is claims
1. "Italian city dump OpenOffice for Microsoft After Four years"
Quotes from the text:
1.a) "we decided we had to keep a hybrid solution, using the two systems at the same time."
1.b) "Between 2011 and 2014, the municipality of Pesaro, in the Marche region, trained up its 500 employees to use OpenOffice, " (sentence continue to c)
1.c) "however, last year the organization decided to switch back to Microsoft and use its cloud productivity suite Office 365."
From the above we can clearly see the headline is biased, The original Microsoft Office package was also dropped alongside Open Office, and all this was likely part of a completely new deployment since the 2011-2014 time-frame indicate the baseline was Microsoft Windows 7 which had mainstream support only to January 13, 2015.
A more unbiased headline would have been something like: "Italian city decide to migrate from hybrid Microsoft Office and Open Office to a new web based Microsoft Office solution".
Next topic - they did choose to use Microsoft Office 365 rather than move to for example LibreOffice during the current deployment apparently due to an evaluation of Total Cost of Ownership (TCO), but then the question is how that TCO was calculated?
It should already be apparent to everyone that the TCO cost of the previous solution was due to their "Hybrid" problems. This cost would have disappeared regardless of choice as long as they did not keep a hybrid approach. And does it list the cost of having a web-based solution - there is not even any mention of potential downtime due to
1) no local access to internet
2) failures of internet providers
3) failures of the external service provider
All 3 appear to cause 100% downtime for the 500 employees in question compared with local install. Assuming 1% downtime and 99% up-time, and a 8 hour day, it might represent 4.6 minutes a day for each worker, or 4.79 full workdays each workday - is it more reasonable to assume zero downtime like the article that talk about TCO?
And what about additional lag time of constant work with web-app compared with locally installed software, - x milliseconds lost every minute times number of employees?
I would not actually be surprised a local install of Microsoft Office 2013 would have a lower TCO than Microsoft Office 365 in a lot of companies despite the higher licensing cost.