Personally I am open to ALL xTalks, including retro-xTalks with bare-metal or via emulation.
I DO currently have macs that can run macOS 6 - 7, 7 - 9, 10.0 - 12.5 (current), but I wasn't planing on trying to back-port anything to do with the IDE.
As far as OXT IDE based on LC CE is concerned 10.9 is the build target for 9.x, which is a real pain in the ass because you need like 4 copies of X-code for the SDKs to build the thing on newer macOS versions. I'm supporting additional (and newer than 10.9) macOS features such as darkMode via Extension Builder. I suspect most things in v.9.xx would work just fine on 10.7 Lion or 10.6-Snow running in 64bit mode, but you can't just change the app's minimum system version in the info.plist (I tried) and have it launch, it would need to be recompiled to run on older stuff.
LC CE 6.6 or 6.7 was last builds with PowerPC support.
Not sure what was last Unix/BSD support version?
I wonder how hard that would be to port the IDE Engine to FreeBSD?
What could probably be incorporated into OXT fairly easily, would be adding the 6.x -7.x-8.x engines as deployment options, and there could be checks to ensure a stack doesn't use newer features before building as standalone. So. then if you have a simple stack with no widgets or SVG graphics, and you want to build it as an app to run on macOS 10.4 Tiger PowerPC running on a 1998 PowerMac9600 (I've done it) with G4 upgrade, you could. I don't think that's worth much effort though, since you could just directly fire-up old versions of xTalks on those systems already.