Originally Posted by Fat Boy

A) Testing frame materials alone would be impractical to the point of impossible and/or meaningless.
B) You can characterize a huge number of ride characteristics with this type of analysis if you want.
C) It wouldn't be all that expensive (<$10k), especially for a manufacturer.
D) Cars are much, much more difficult to characterize and it's done every day.
E) Car chassis construction plays a huge part in ride characteristics and NVH concerns even with a suspension. Modal analysis of a car chassis is a big part of the design and development of any car.
F) Tires are a huge component. The compound is less important than the construction of the carcass, but they're both players.
G) Rider impressions aren't perfect, but they're a good starting point and when correlated with numerical data can be surprisingly accurate in a qualitative sense.
H) Any manufacturer that isn't doing this isn't serious about building a good bike. Having said that, I don't think it's common practice.