I bet this has been asked before and I just didn't see it... anyway, I'm wondering if that was intentional that TAB visibility cycling doesn't do anything if the insertion point is after the ellipses indicating folded content.
** Example header... ^1 ^2 ^3 ^4 If the point is at locations 1, 2 or 3, TAB will reveal the next level of children. (The ^ locations will make sense if you format the e-mail using a monospace font. A proportional font will just look nonsensical.) ** Example header *** Child node If it's at location 4, TAB cycling does nothing. It's a minor point, but somehow when I'm moving the pointer around in Emacs, quite often it ends up at the end of a displayed line, and then I have to remember to take an extra step to make sure TAB does what it's supposed to do. I'll grant, there could be a good reason for the behavior, but I think it's reasonable to expect TAB to expand a heading if the point is anywhere on the line (including at the end). hjh