This seems to be the physical bounds of the image volume. ImageStacks expects you to enter the correct scaling manually. That 1.0x1.0x1.0 is just simply a place holder (it is a common resolution for medical imaging).
So 5 micron (or rather 0.005 as units in Slicer are all mms) seems reasonable given the ROI bounds and the number of slices you reported. But you still need to find out what the resolution in X and Y fields are. In confocal they tend to be much higher resolution than slice spacing. If your metadata does not show you that, then you might be able to get a close enough approximation by dividing the X and Y lengths reported to the X Y image dimension (in pixels). For example if your images are 1024x1024, dividing 1612/1024 will give you the resolution of 1.57 micron/px. Then you can enter this to the XY fields of spacing as 0.00157.
I am not sure what tutorial that would be. Official SlicerMorph tutorials are here:
Specifically this is the link to the ImageStacks tutorial with some toy microCT data