'*My* MTF talk', lol. Authorities describe optical bench MTF as 'an accurate characterisation of lens performance'. Not the full story, of course, no one paints it as such, but a deadly accurate test with respect to what is being measured. I read road tests and specifications for cars before buying too.
Aham, I don't buy lenses that do not have what I want in performance for *my purposes*, not enough time to buy and try and sell if not up to scratch. This is one lens I would not buy for several reasons, outlined above, but I don't shoot urban or street or interiors or shows with motorcycles. I am sure it's a fine optic, that is its reputation.
The mid-frame performance drop-off of Leica R wides is well known, and, like the excellent centres, is a design intent to produce other image characteristics. Image content data, colour and processing play heavy roles in obscuring lens performance, so you may not see the differences between lenses in small downsized web images. Don't believe for a moment they do not exist however, unless of course you care not to respond to factual data, and that does have its appeal to some people, of course.
Andrew, the linked images are a poor test of a lens's performance: close focus, no graduated fine detail, coarse structures, and so dof and image content play a huge role.
For a good recent example of how to assess fine detail and lens performance in the real world, see johahill's 35mm lens comparo - last few days from now. It should be methodical and well documented, and use best practice capture methods - and suitable subject matter. Archives probably hold a lot of such material also, by denoir and others.
But - if owning and using is believing, fair enough, each to his own.
