I know it’s quite an old post but maybe there is some update in that subject. As @Rahmi_Ozdemir pointed above, layout marks printed onto PCB, leads to conclusion that it could be possible to desolder old socket and mount one, that iclude another GND and 12V line. This could introduce compatibility to 3.5 inch drives. I’ve stumbled upon some photos of R2 (old revolutionary board?) that actualy had that bigger socket. Was it working back than for 3.5 drives?
I’ve looked into my HDD spec sheet (WD10EFRX) and as I saw, it consumes 1,2A at peak. Couldn’t find precise data about R2’s average power consumtion but basead on informations about recommended power supply (12V, 2A), I assume that something with 3,5-4A should be fine. May be it would do the trick? Unfortuantly, for now I’m using 3A power supply for R2, so it might not be enought
I’ve replaced the default 2-pin socket with 4-pin, and use it with 2.5" WD VelociRaptor WD1000CHTZ-0. (12V current is rated to 0.35A, IDK about spin-up peak current). My R2 works stable with a 2A power supply, including spin-up on power-on. Also, HDD remains powered-off with DC plugged-in, and R2 is powered off.
So your case is pretty much simmilar to mine. As You pointed out, after connecting power supply, the 12V line is immediatly feeding that 12V to the pins in socket but 5V line is not. I checked it with voltmeter and it starts delivering 5V to the pins after the boot process (after few seconds of pressing power button). I’m curious how it affects a drive, when it has 12V from the begining and gets 5V after some time. Have you done someting like bridging power button or does it work just like that, with few seconds of delay?
As I know 12V is only for mechanics when 5v is for MCU/logic (correct me if I wrong). So in case when mechanics is controlled by logic it should be ok for a HDD, when it has permanent 12V, anyway it starts/spins up only when it has 5V, which is controlled by SoC’s power button.
P.S. It’s only my opinion that might be wrong, so all modifications are at your own risk