About your point 8):
I have had the same experience many times. The Surface seems to be rather picky as to what input method wakes it up. Sometimes, when only the screen has been turned off, opening the cover is enough. If it is asleep, you need to press some buttons on the
keyboard or mess around on the track pad. Other times, it only wakes with the power button.
I have a lot 1544, which miraculously lost the sleep of death problem (after it first occurred in January) in the middle of February (before the updates were supposed to fix it).
The Surface is configured to sleep after 5 (10) minutes when on battery (plugged in) and to never hibernate, so I think if there were any further sleep of death problem on my Surface I should be facing it all the time. Connected standby is on in both cases
and Windows Hello is enabled. I figured there is no use to an expensive laptop if I disable all of its cool features.
Don't accept disabling features as a fix!! (This goes out to the recent comment about disabling everything!)
For the majority of wakes everything worked fine. Once I had a reoccurrence of the sleep of death (at least the symptoms), but after letting it sit for about a minute or so it came on.
Ironically, just as I was writing this I had another failure to start. I began writing this post in the office, then closed the cover, put the Surface in its sleeve and took it home. Here, I connected it to the power brick as always and then opened the cover.
Pressing buttons did light up the keyboard, but not the screen. Now, I remember that the sleep of death would have a lit black screen, i.e. the backlight is on, but the screen doesn't show anything. Today, the backlight didn't come on and also pressing the
power button - repeatedly - didn't bring it to life. After trying for a couple of minutes I gave up and forced a restart, only to discover from the event log that, apparently, two seconds before I closed the cover there were a bunch of errors ("An attempt
to configure the input mode of a multitouch device failed."). These are listed before the system entered connected standby, which is the last entry before the unexpected restart an hour later.
I wouldn't bet that these are somehow related, but what I just experienced (for the first time) looks a bit different from the usual sleep of death, where the screen wakes but doesn't show anything. I'm hoping this was a one time thing and won't happen again...