Hi All,
Trying to put forward a completely different point of view:
I am still not convinced that the UCV is the right approach. Here we are forcefully trying to fit a union by creating the single dimension table. It can be acceptable in the car dealer scenario since the example tables are small, but in real life practical scenarios with large tables, such single dimension table might be as big as the other fact tables.
From the purely database perspective, I would expect HANA to allow me to perform table joins with the acceptable performance. If the performance is slow, then SAP should fix it. Unfortunately we try to create workarounds for simple business requirements and spend quite much time and effort to maintain such workarounds. At times it gets hard to convince the business users why their simple requirements cannot be met due to limitations of the technology.
So, I think Rajarshi should raise OSS note, take this example to SAP and let them figure out a way to optimize the HANA engine to allow large table joins. May be, SAP will make it available in SPS08 or 09 
Regards,
Ravi