Without test equipment and laboratory conditions this is purely a subjective single person view point.
How will you test for the following which because they happen to work in a Saab doesn’t work at all in another set
Geometric characteristics of the lamp
5 Axis reference of the filament
Quality of components. Quartz or hard glass, Nickel plated brass or tin
Life of the lamp measured against STD Weibull procedures
Out gassing
Wattage (no the figures on the base have no relevance to reality)
Tungsten evaporation process and Lumen maintenance curve
There are a number of reference points in the beam. Is it good to have a very broight single pooint or a good spread. How will you measure it?
etc. etc.
Unfortunately vehicle lighting is a hugely complicated and complex series of solutions. This is precisely why you pay double for one lamp compared to another.
While I respect your initiative and reasoning it will be costly and inconclusive and quite honestly meaningless except to you, in your car on any given situation (Is it raining or is it clear for example)
This has been done at a basic level over the last few years in a number of independent test in magazines and this is a start with a far higher degree of accuracy
Nick
http://www.hid-online.com/hidonline/