I've had Norton Safe Email for several months now protecting one GMail account and two Yahoo accounts. The annual subscription is modest right now at $19.99/year but it hasn't actually delivered much value yet. I have hope that as the service matures it will.
The concept of a cloud-based email protection service is a solid one as it can really be a value to individuals that access email accounts both through apps and web browsers. I implemented a competitor's cloud-based service a number of years ago in the county healthcare system I was a part of and the numbers of user issues with email delivered malware, phishing and other exploits went way down. Norton's service hasn't been nearly as effective.
Norton is very clear that they can only monitor email accounts that support POP3 and SMTP and given that this is a consumer-oriented product that makes sense. For the email vendors they support, this generally involves Norton adding a clear label to an email indicating that Norton considers it Safe or Suspicious (as in GMail) or it creates a folder/sub-directory called "Norton Suspicious" where it moves all messages that it considers questionable. The basic design is simple and fairly easy to understand but in operation Norton currently falls far short of other cloud-based email protection services (granted, most of those are aimed at commercial and government accounts).
All email filtering services generate some number of false positives (flagging a good email as suspicious) and false negatives (where the service lets a bad email through as good) and this is where Norton is currently not doing an acceptable job. Customers can send samples of bad emails that Norton is letting through as OK but the timeline to their action on this information seems to be weeks (if at all) instead of the few days it needs to be given the current threat environment via email.
The flagging of a good email as suspicious is even worse from my experience. Norton Safe EMail makes no provision to let an individual customer set up a domain whitelist for themselves (to specify communications they always wish to accept) and if you manually move such an email out of the Norton Suspicious folder back to your inbox within a day Norton Safe EMail will have moved it back into the suspicious folder. I've never seen where Norton changes their algorithm that a given suspicious email is now considered safe after receiving customer input. I can understand that as that could give an avenue to the bad guys to get their messages to be considered OK.
If a service like GMail or Yahoo flags a message as spam and puts that in the OEM's spam folder, Norton takes no further action and that is also understandable. In roughly 6 months of use, I've never had Norton Safe EMail flag a message as suspicious in GMail although GMail has marked about 20 as spam (and legitimately so) during that time. For Yahoo, Norton Safe Email has flagged 16 messages as "suspicious" but only 2 were legitimately so - the balance were false positives. It also missed eight messages that were clearly phishing attempts with broken English and/or embedded URLs clearly not matching up to the stated sender.
Norton has been getting better at not passing on such phishing attempts recently.
I do think that Norton needs to provide each subscriber the ability to specify a domain whitelist, however, to eliminate the hassle associated with the false positives it generates.
My subscription renews in May, 2025. Hopefully, Norton will make sufficient improvements to make it worth keeping.