Roger's testing is interesting especially because he often shows sample variability in the measured MTF.
I can also see the motivation for the take-apart pieces. He needs to understand the construction of lenses to be able to identify problems in rented gear and how best to rectify them. However, I don't think evaluating the durability of something just by looking at it and taking it apart is going to give the full story; only time can tell which solutions are durable and rugged over long periods of time. Thus I would not read too much into this.
It is embarressing that early information from Nikon suggested a ring SWM however, what matters is if the shots are in focus or not, and in this respect the 105/1.4 excels. Roger notes "The care taken to engineer a smooth, accurate focusing feel is very evident. " " Its focusing system is still excellent and accurate, no matter what kind of motor is driving it. " "Nikon lenses are just as reliable as anyone else’s."
Although it is not an AF speed demon, the in-focus rate with the D5 has been excellent.
If I have any complaint about the lens, it's that I would have preferred a bit more modest f/2 version for compactness and probably the swirly bokeh could have been avoided. However, the consistently high quality of the images makes my complaint not so severe. I do like the lens, although I think it's a bit excessive on the specifications. I do take advantage of the f/1.4 aperture and it lets me often shoot at ISO 2000 to 3200 instead of 4000 to 6400 or worse, in indoor available light (and outdoor night light) situations. Because it's so well corrected, and because of the high accuracy of the autofocus, the images can be cropped and they are still sharp enough so that I don't notice that I have in fact cropped the images, in the final results. Thus I may be able to use the 105/1.4 in place of a 70-200 when the light is so low that this is advantageous. However, for outdoor travel use I think a 105/2 would have been nicer.
When my 24-70/2.8 G (uneven stiffness of the zoom, a mark on front element, due to dropping) and 70-200/4 G (faulty VR system) needed repairs, the solution proposed in both cases was to replace the entire inside lens barrel including optics, focus motor etc. and insert it as one piece into the old lens's outer chassis. This has got to be really expensive for Nikon. I can understand that in a complex zoom lens, this may be the most practical solution, but I would really think they need to think about how to replace individual faulty components on the lens without replacing everything. But they must understand the economics of repair much better than I do.
