I believe this is so.
Need to take into account the difference between the diagonal on the 35mm film (24mm x 36mm --> 43.3mm compared with the diagonal on the camera chip (22.2mm x14.8mm --> diagonal = 26.7mm) ==> 43.3/26.7 = 1.62 ==> for a lens designed for 35mm film , the effective focal length is 1.62x longer.
BUT if your MD - EF adapter is like mine, because the MD registry distance is different to the Canon EF registry distance (
http://www.photoethnography.com/Clas...html~mainFrame ), there is a corrector lens built into the adapter to move the focal plain to the correct registry distance to permit focus at infinity and you need to correct for that (I have been told the correction is multiply both the effective focal length the and the f stop being used by 1.4x.(Typically the suppliers say the corrector lens factor is 1.4x for an MD-EF adapter.)
So I believe in effect you need to use a factor of 1.6x1.4 =2.27x for the focal length on a 20D chip and 1.6x for the f.stop.
I think I have that right. ?
So f3.5 become f3.5x 1.4 = f4.9 ?
and 200mm MD on the EF become 200mmx 2.27 = 454mm Eff focal length. ?
If the 200mm Tamron had an EF bayonet and was designed for Canon EF cameras and no adapter was needed to use it, the correction is just 1.62x to get the effective focal length and I think the f number says the same.
If anyone thinks these calcs are wrong , feel free to say so.