I like the flexibility that IMAP provides in allowing me to access a Master Source via multiple devices; each of which updates and reflects the single state once synched.
However, I also recognise that this is a Single Point of Failure, so mitigate that risk by taking periodic backups, stored in multiple places, locally and remotely. A good rule to have is: never (ever) rely on other people, organisations, equipment, hardware, systems, etc., to look after the data you hold dear. Always build in some redundancy suitable to your level of acceptable risk.
Additionally, I regularly (probably daily) synchronise each device with the Server which, in this particular instance, highlighted the issue with one device, allowing me to take decisive action with the other devices – i.e. backing up the data on them and deferring accessing the server until the issue has been resolved. As such, I have a recent backup from my regular process, together with more up-to-date versions on the, currently email-dormant devices.
All that said, I am fairly tech-savvy, aware of the client-server architecture of IMAP and, thus, appreciate the inherent risks of 'relying' on that server-based source of master data. Reading the messages of the many frustrated peeps on this board makes me think that some thought the data on their device(s) was the master copy. That's through no fault of their own; perhaps organisations like VM, providing such services, should be much more forthright in highlighting such risks as well as suggesting ways to mitigate them. At the end of the day it helps both customers and organisations to be so proactive...