It looks like there are a lot of reasonably inexpensive DROs out there.

X and Y are mass-sensitive, have to move fast, and don't generally need to be super-repeatable when it comes to homing, so limit switches make lots of sense and DROs wouldn't.

Z is typically not so mass-sensitive, moves slow, but should be very repeatable for homing, especially with non-sprung beds where a head crash is extra bad.

Many DROs preserve absolute position across a power-off, and many of them speak known protocols.

I did google and site searches, but I don't see any instances of using a DRO for closed-loop Z absolute position. I would still want a limit switch as a backup, but setting an absolute origin in a DRO and then going straight to the first-layer Z height without dancing on a limit switch feels like it could be a win on a sufficiently stable machine.

Just pondering here...

Am I missing something obvious?

A snap action switch provides precise enough Z=0 (or Z=1, if you prefer to use an offset for safety), assuming you use a quality switch (about $1 instead of $0.10) and take off any levers.

Springs on bed leveling screws aren't there to protect against head crashes. If they could do that the bed would be bouncing around all the time.

Head crashes aren't a matter of getting a more accurate position sensor. It's about firmware reliability. If the firmware is set up incorrectly or ignores the switch/sensor input, the nozzle is going to crash into the bed whether you use a switch, a hall effect sensor, an opto interruptor, or a DRO.

Better documentation on properly configuring the Z axis would help prevent user errors.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]