In a nutshell, I provide two (actually more) kinds of contact avenues depending on which problems real people may encounter and why.
Note: The following scheme sounds more complicated than it is (& Jim Morgan would probably think it insane:)
1.) My on-site redirected-to pages provide info and show echo'd details about the visitor's Host, App, etc., and also describe which details to send to the e-mail address displayed on the page as a graphic.
2.) My 403s get the worst of the worst so they have mucho hoops --
- My standard 403 provides the same info but includes a link to an off-site page -- a bare IP I use as a catch-all for troublesome visitors redirected from active domains -- that then shows the same echo'd details.
- But rather than a graphic e-mail address, the off-site page includes a link that pops up a window with a reCAPTCHA Mailhide. [google.com...]
If I didn't have a catch-all IP, I'd use #1 for Custom 403s, too, and possibly include the Mailhide pop-up. But the graphic address is enough, really, and shows in your logs that a real browser at least hit it.