I also think that, historically, electronics were documented because they broke down so often.
You just had to be able to spot a broken tube in your radio, clean a potentiometer, etc.
There are only so many way ways to skin a cat (in nuts and bolts engineering). When constrained by physics, everyone is going to converge on the same fundamental designs. You therefore have two or three options. Take or leave the licensing agreement of the IP holder, which need not be reasonable or effective; say screw it, use the IP and hide the schemata behind enough trade secret bs red tape that no one is comfortable leaking it, or give up.
I wager the second option contributes to things when number 1 fails. At least, it's the only reason that holds water for me, then again, I haven't gotten into that industry enough to vouch for the presence of skeletal remains in the closet.
One way to look at it is as ('at it is as'? English is weird) general purpose computer vs. appliance. Apple probably feels no more obligated to give you schematics, chip internals, etc. than Garmin does for a GPS.
Another angle is that hardware companies are becoming more secretive over time. There weren't many secrets in a Marantz receiver or a shortwave radio. Perhaps there's been a change in approach due to low-cost Asian manufacture and the tendency to ignore IP.