I'm sure someone asked for the functionality the way they've set it up in CREO 4.0, but restricting users from being able to attach datum targets to datum planes and axes is horribly short-sighted.
I realize it makes things a little faster for the initial drawing creation or MBD, but I'm guessing this is going to create some headaches as parts, assemblies and drawings go through a few revisions.
Consider something with a multitude of parts, such as a weldment. Previously, you would select a plane as a datum for something like flatness. All your mating components are assembled such that they have a face coincident to that plane. By selecting the plane rather than a face, your plane transcends any geometrical changes in the individual components.
Even when producing detailed drawings, you may want to physically move the datum target outside the edge of the physical geometry. Well, guess what...You can't. Thanks PTC.
Just create points on your datum plane and attach your datum target to those points instead of the plane as a whole. They will stay with your plane regardless of component geometry changes. It is almost trivial to create a point in plane with a flat surface but beyond its edge. You also gain the added benefit of increased precision in placement.
It tends to be better GD&T thinking to have your datum planes based on 3-2-1 points anyways.
Rename the plane in the Datum tag what you need or create another plane and name it with the needed Datum Tag. This will work for planes and axis. The only one issue what I found is you can't manipulate the location of the axis datum tag anymore in the drawing environment. The planes you can still manipulate.
That is exactly the method used in CREO 3.0.
My issue is that now, when you rename and attempt to show the datum target, CREO uses the 1982 version of ASME Y14.5 for Datum targets (not the current with the perpindicular leader and filled triangle) and you also no longer can associate GTOL features with them.
This forces you to use the physical surface as the reference.
Perfect scenario: Weldment with multiple planar faces with the intention of them creating a datum plane with a combined planar tolerance. By selecting a face of a part rather than a plane, what happens when you revise the weldment and the part the datum is associated with is replaced/changes/moves? Well, your datum moves also. Had you been able to associate it to a plane, the individual faces are not critical, because what you intended to be your datum, stays your datum.
I will check this method out. I do understand your point about the thinking behind it, but it's really only valid from the application of real-world GD&T. In a model, additional geometry like this only increases the risks of instability.
With the "improvements" intending to speed up the process, adding points only defeats that goal...
Thanks for the suggestion though! That sounds like a reasonable workaround.