And what is wrong with having a feature where the email address is actually checked before the email is accepted? A do_not_reply email address would actually be valid and registered even though someone would not read the emails sent to this address.
No, many "do not reply" email addresses are
not valid, i.e. if you send a message to them, they will bounce. That is the point of the "anti-spam" measure, i.e. if it bounces, a robot is less likely to pick it up and move on. Of course, "robots" have been "getting smarter" with these now...
Some "do not reply" email addresses
may be valid, but otherwise would just be a repository which is not checked. It's just more time and cost efficient to have a bouncing address. It also gives recipients the "right message" (i.e. legitimate ones), i.e. if they try to send a reply and get a bounce, then they're more likely to realise they've done the wrong thing (i.e. go find the correct email to send replies / feedback). If they reply and instead get accepted to a "black hole" of an address, they get no response and will be confused when their message doesn't get answered for a long time.
The fake email address(es) used by Yevgeny on the other hand would not be valid and if they were then I would be spamming the hell out of them.
As I said, some are valid and some are not, so some will bounce, but others will not. Nigerian 401 scam emails are typically responsive, which is of course the vehicle by which they start working on their victims.
And I may not be the same as you but I see people sending email to my junk mail as an invasion of my privacy. Especially the same rubbish over and over. Perhaps I am asking for too much in this day and age where anything is possible.
The best you can do is filter spam straight to your trash on a perma-delete. Of course, you'll still have the problem of the odd legitimate email which will fall into the trap, and in the latter case you won't even know it until the sender perhaps prompts you in another manner.
Spam is getting smarter, unfortunately, and the task of filtering the good from the bad is both subjective and, as hard as it is to believe, a non-trivial undertaking. For example, I keep getting spam in Hotmail about Facebook friends inviting me. Facebook itself is legitimate (and let's not debate that point here) and so is inviting friends, but these invitations I'm getting are obviously spoofs.
Actually that gives me a thought. If only I was able to send spam emails.
I'm sure there must be enough underground sites out there that could teach you how to set up a spam robot.
Of course, make sure you don't get caught - there
are laws against spamming and there are notable penalties. Unfortunately, there aren't a lot of resources out there dedicated to catching spammers, and spammers are already and increasingly notoriously difficult to catch anyway (and let's not even start to get into establishing a burden of proof).