There are micrographs to support both arrangements of SP-D fuzzyballs.

There is side to side N termini attachments for SP-A multimers, why not SP-D, since clearly there are images which support this arrangement.  The publication by Arroyo et al from which this fuzzyball is taken has a different LUT plot than the one i see.  I see a definite dip in the lightness (brightness, luminance whatever you want to call the grayscale) in the path of the two arms they created. (Their diagram is below – with my comments below my crop of their table of LUT plots).

My red arrow  (left, second image from left) lies right over an area of low greyscale values (you can also see this on my rendition of their AFM above) but their plot to the right of that image (shows the peak grayscale (that is the higher height value on their Y axis) on the wrong side of the image.  Two bright bumps in the plot, on the left, but in the image they are on the right.  Second problem with this portion of their diagram is that the fuzzy – ball they draw has NO central denser area. That makes it more equivalent to the fuzzyball in my group of images (labeled 1) and the diagram to the left of it.  Whether the solid center height (lightest of the grayscale (highest number and greatest height) could be the glycosylation site and N terminal or might just be the N terminal in a V shape with side to side attachments.

My personal view is that the three collagen-like domain peaks (which could include some of the coiled coil neck domain) remain as clearly visible peaks along the arms of the fuzzyball, though they are a little difficult to plot.

The next step is to measure from the very same micrograph the LUT plots and arm lengths of trimers, hexamers, dodecamers and fuzzyballs to see whether there is a significant difference in the diameter/length that might account for the dark center of some SP-D multimers. And I am not saying anything bad about the measurements of Arroyo et al, just want to see whether there are other things to be gleaned from their very nice AFM images.