I agree that this problem appears to affect any Surface Pro 4 or Surface Book - these machines are simply incapable of properly managing their power states, especially when used with a dock and the keyboard cover.
My SP4 now seems to wake successfully *once* now after undocking and sleeping, but I think the next sleep cycle is doomed to fail, regardless of whether it's triggered by pressing the power button, putting the SP4 to sleep via the Windows Menu Power controls,
or simply closing the keyboard cover.
Yesterday, as I said, it woke successfully the first time, just by opening the keyboard cover, but then I put it to sleep to go to a second meeting, and it was nearly hot enough to cook an egg on when I took it out of its sleeve a half hour later. This
is only about the third time I've encountered the "hot bag" problem, but there's no way that that amount of heat didn't take many months of life off the battery of my seventh Surface device (5 ea SP3's, including 3 obviously defective refurbs, and now 2 ea
SP4s...) All of my settings are default EXCEPT Wifi Connected standby is turned off on both battery and when plugged in, and it is set to never sleep when plugged in. (Power, sleep buttons and lid closing are all set to "Sleep".)
Sooner or later, Microsoft is going to start a fire with one of these things, which may explain their total and complete silence on these issues - acknowledging the existence of the problem in any way could really blow back on them in any legal proceedings
(which at this point seem to be a matter of if, not when...)
I will say that the discovery for whatever lawsuits that come will be extremely revealing, as it appears that Intel and Microsoft (and perhaps Pegatron) are colluding to cover up what must by now be a pretty well-known (and understood) problem with the Skylake-based
SP4 and SB within their respective companies. My guess is that there's really just no way to fix this in firmware and/or software (requiring total hardware replacement, which is obviously quite expensive!), and they're arguing between themselves about who's
going to pay for replacing every single SP4 and SB out in the wild, since they're all clearly faulty and potentially dangerous...
That said, it seems possible that there may still be some way to sort out the race conditions and state machine blocks and collisions (dock/undock, different ways of putting to sleep, keyboard cover click-on/off, keyboard cover opening/closing, power state
transitions, bluetooth/wifi events, etc....) that seem to be at the root of the problems by putting all power management control in *one* place. I presume that place is the firmware, and that this will require that these machines (all Skylake machines? Just
the SP4/SB? Who knows?...) to interface differently with Win10 power management APIs than all other PCs. If this is not possible, then we, as the owners that got suckered into buying these turkeys, seem to be stuck with machines that can never be made to
work reliably and safely as designed.
It would be really nice if Microsoft would step up and tell its customers the truth about what is going on (the "really hard computer science problem involving new hardware"), but it's pretty clear the lawyers won't let them do that, so we continue to suffer
and not get what we paid for while they *may* continue to work on a fix, if one is even possible...
I love my SP4 and want it to work, but I've lost all faith in Microsoft to deliver a reliable product.