I mean I know the technical reasons. But it says on the FairEmail page: "FairEmail might be for you if you value your privacy." and also says "works with virtually all email providers, including Gmail, Outlook and Yahoo!"
For me, privacy is more than what's on the client - the server matters too. POP has seemed to work for me for 25+ yrs.
You will never escape having some of your email stored on Gmail servers. Most people you email are using a gmail address, whether direct or via forwarding.
End-to-end encryption is your only true ally. In this regard, email is hopeless. I personally feel that all these privacy mechanisms on top of email are a) hopelessly pointless and b) give people false comfort. They do more harm than good. It would be better if people saw email as a plaintext, insecure protocol and treated it that way at all times. Just imagine whatever you write in an email is the same as a message you send out to the world on Twitter. Doing it this way ensures you never send a message you will regret and will guarantee your protection rather than the security theater of privacy mechanisms layered on top of a fundamentally broken protocol.
Also, I'm not sure most email users know what a "plaintext, insecure protocol" is and what it would imply.
If both your email provider and your client (aka mail reading software) support it, there really is no reason to use POP instead of IMAP.
While I agree that POP still uses the server as a go-between, at least the mail doesn't reside on POP servers forever. Whereas with IMAP, if I have 25+ years of email I'd like to be able to view and archive and search, all of that has to sit at the server rather than at the client.
That's very dependent on the POP server. The protocol only tells the server that it is allowed to delete the message, not that it must.
I'm pretty sure if you use POP on gmail it just does an "archive" on the backend and the mail is still there, for example.
getmail supports deletion as an IMAP client, for example.
I already trust my provider to handle my email once; there doesn't seem to be a vastly larger trust requirement on my part for them to handle it multiple times. (And epsilon additional privacy concern.)
Fair enough, but that strikes me as a very odd risk posture to take. Either your email privacy is valuable to you, in which case I would think you would want to protect it against non-lazy people as well, or it isn't, in which case what difference does it make?
But it's obviously your call.
A hack will barely get them anything, neither a warrant or a bored employee.
Of course any of those situations could result in an active tap that stores everything. But that is orders of magnitude more effort and still doesn't get any history.
Just minimizing the attack surface.