My most recent one was Apple telling me my id was reset by my request. This definitely came from Apple, was verifiable... But it turned out some person entered my email on their support case and it almost gave me access to their entire Apple account, the support people were willing to go through everything with me as I had the email.
It was only when I asked for the transcript of what I'd apparently said to them that alarm bells rang on their end and they finally investigated enough and escalated enough and determined my email was not the one that had anything to do with the account in question.
Mostly I ignore the email I receive, but once in a while it has enough details that I can find the person involved and give them their train tickets, or car insurance docs, etc.
It's interesting to think about how the very basis of our internet identities, that is the email, can be so easily abused by someone who has bad intent.
What's wrong with "Click the link in the email we just sent to [email protected] to verify your email"?
But it could be mitigated by "click the link and enter the one time code we gave you at sign-up time". Too much friction? How about "click the link on the same browser you used to sign up, and we'll verify that using a cookie we just set" - functionally equivalent and probably works for 90% of users while the rest can fall back to the one time code.
I've seen a handful of sites do something like this in practice. No idea why it's not more common: presumably most people don't roll their own verification process so if some major web frameworks adopt it we'll eventually see it more widely.
Oh boy. Auth is that thing that looks so easy because you just need to store an md5 password to feel like hackerman. If people actually used existing solutions, web logins wouldn’t be in such dire conditions.
Not all careless stupidity should be attributed to the website admin, however: assholes using random email addresses and phone numbers deserve to be punished, and knowing one's own email addresses is a basic literacy requirement.
I have a very common name and a very common surname and people have used my email (name.surname at gmail.com plus the infamous GMail variants such as namesurname or NameSurname) for purposes like accounts on dating sites, Spotify, Instagram etc.; invoices; banks and insurances; resumes and job applications; medical test results; newsletters and all kinds of personal communications.
"I" am a local politician, a Swiss or Italian banker, a boyfriend deserting some girl in Argentina, a rugby player, a professor or two; "I" buy screws, magic tricks, diving suits; I know several of "my" birth dates and addresses and I have easily identified a couple of correspondents.
In most cases there is no practical way to verify email addresses, particularly if the person is really convinced that their email is the wrong one or that some approximation is allowed, and without actual payment collections coming your way little harm is done.
I sometimes complain to web sites with inexcusable confirmation-less registrations or reclaim accounts on services I might want to use, but for the most part I just let incorrect emails accumulate to play the passive game of collating them and consolidating identities (e.g. is the person who follows cooking courses in a certain big city the same who received from a friend bus timetables for that city?).