IMac – USB3 HDs fail to mount on wake from sleep with new iMac Pro

Ideally, you should set a longer spin-down time so that it is not occuring constantly. A time to spin down of after being idle for 4-6 hours would be better. This gives you the benefit of saving power for long periods when little is being done on the drives, e.g. during the night or while you are away from home at work, etc. Having longer times also allows for less slow down when you are using your drive since you will have a reasonable time-frame to get files and work done before the drive trys to spins down on you.

Use pmset from the terminal to set up a custom interval that works best for you.

Others our claiming that spin down or not is really not that big of a deal in a home / SMB environment.

As for the differences between the 3.5" and 2.5" drives its safe to assume they are the same or similar. Even a comparison between what you said was a 3.5" server drives is not the same as a comparison between a 3.5" consumer desktop drive, they are not on equal ground just because of form factor. It really comes down to the manufacture quality and is really one of those you get what you pay for kind of situations and OEM drives even in Macs are still a lowest bidder kind or process. Here is a discussion on the reliability of 2.5" vs. 3.5" drives.

As for the differences in trying to set internal or external drives independently it looks like its not possible since the setting is really only intended for the startup disk according to The X Lab -- Hard Disk Sleep

Another undocumented aspect of Put the hard disk(s) to sleep when possible is that this option generally applies only to the startup disk. Non-startup hard disk drives — external FireWire® or USB® drives or additional drives installed in a Mac desktop — may ignore this setting.

I sometimes have this problem with my built-in keyboard (and trackpad). It seems to be related to static electricity build-up (in me). In dry environments prone to static generation (e.g., carpeted offices, fabric-covered chairs, etc.), I always try to ground myself on something (say, a filing cabinet or a metal wall) after I sit down but before I touch the mac. If I forget and I zap the keyboard when I start to type, there's a very good chance that the keyboard and trackpad will be unresponsive. Based on my recovery action, I'm guessing that the macbook's internal USB controller has sensed the transient voltage spike as an over-current condition and has shut down (suspended) the offending devices on the associated port.

To recover the keyboard and trackpad, I either plug in an external keyboard (or vnc in via an ssh tunnel) and run the USB Prober app (part of the Developers tools, under /Developer/Applications/Utilities). Inside Prober with the Bus Probe tab selected, I then check Probe suspended devices and hit the Refresh button; my built-in keyboard and trackpad immediately start working again. My assumption is that the probe causes the suspended devices to re-negotiate their bus membership --which, if I remember the USB protocol, means that they are essentially new devices and not subject to the previous suspension -- and lets things go back to normal.

This doesn't happen often enough for me to automate the recovery, but I'm guessing there are associated system events/messages that could be used to drive an applescript-controlled probe.