The big thing is, it's not like Autodesk isn't aware of all this stuff.
But we get a 2009 release that does little to address the issues. I'm glad we can place parcel labels on parcels in XREFs now, but that's really just a patch on the greater problem of not being able to DREF our parcels. And really, we need parcels to be redesigned from the ground up. We don't need more and more stuff built on a faulty foundation.
The more stuff they tie to these errors, the less and less likely it gets that anything will be fixed. Take the terrible Project Management support. We desperately need a way to create a PROJECT in Civil-3D, where we can set all those important project-level parameters like Coordinate Zone, Units, Project-Type settings (e.g., Grid or Localized, and for a Localized Project, the grid/elevation scale factors or combined scale factor and additional translation/rotation parameters to get to Project Coordinates, if applicable), Project identifier information (name, job number, client name, etc.), the CAD Standards (if any) to be used for this project, etc.
What we get instead is absolutely nothing in Civil-3D, and we are told to use the Vault for PM. However, using the Vault to control all that stuff I just mentioned is an extreme error in design. It makes the Vault highly-coupled with Civil-3D, and means that any change in C3D can require a change in Vault, and vice-versa. Now add in the fact that Autodesk is trying to use Vault with other products, such as Revit, where they will undoubtedly be doing similar things. If Autodesk continues that trend, eventually all of their products will use the Vault. And if every product is as tightly-coupled to the Vault as Civil-3D is becoming, then we have a giant nightmare on our hands. Changes to Revit will end up creating bugs for C3D users, and so forth.
And there are quite a few issues like this. Just getting C3D to work properly for Grid-based projects will take a huge amount of work. But every new release of C3D is simply patches and more features pasted onto a bad foundation. And I think we all know what happens to buildings when the foundations are bad... We need to start seeing some fundamental, significant changes to C3D or the whole thing will collapse.
Where we're heading now is that, in five years when Autodesk builds themselves into a nightmare with C3D, they'll go buy some other competitor or third-party app, and then we'll have the whole transition-from-LDD-to-C3D thing all over again, but this time it will be the transition-from-C3D-to-MoreOfTheSameAfterAnotherHugeLearningCurve.