Linux power management for laptops is mostly a hardware and OEM problem. Every system has at least a few components that don't follow spec for power management, and the OEMs ship their workarounds in Windows drivers rather than in the system firmware. And none of those issues are publicly documented, so Linux developers have to reverse-engineer what PM features are actually usable under Windows on a certain system rather than trust what the hardware and firmware declare support for.
I do not disagree, but that’s where the FOSS structural incentives tend to fail. WiFi chipsets have the same “institutional problems”, but everyone needs WiFi so enough hackers will pour over the problem and generally find decent solutions. That’s generally not the case for power management, because most people will be ok with keeping “the brick” connected most of the time. Every once in a while, this or that company will throw a bit of money at the problem and solve it for a few models, for a few months... and then we go back to square 1.
I don't think the lack of appropriate structural incentives has anything to do with FOSS. It's that Microsoft has a near-monopoly on operating systems, so it's easier for PC hardware vendors to work directly with just Microsoft to deal with problems, rather than publicly document their hardware errata for the benefit of multiple OS vendors.
This is no different for any other chipset under the sun, from the good ol' "winmodems" of the '90s to today. But some stuff gets fixed and some doesn't. Power management is one of those that "doesn't".
