Hi, Megan-

Re. #1: I think that is actaully the vstat block you are looking at and referring to (not the va2t) here, right? I will assume so in replying...
And indeed, the patterns observed here will vary depending on the task/stimulus and modeling. In the example you refer to, there was a very strong audio and visual task presented as a block design with many blocks---we would expect large F-stat then in the temporal lobe and visual cortex, because of both the nature of the task and the relatively simple block design (20 s blocks). In more modern cases, particularly with event related tasks and more complicated designs with fewer regressors per stim class we might expect lower F-stats in many cases. There is some notable visual cortex with high F-stat in your image though, I believe?
And yes, in many cases, you might also expect to see more regions outside the brain, again depending on the task and paradigm. Seeing the eyeballs can also be common, as blinking might be associated with visual stimulus changes.

Re. #2) Hmm, I wonder if there is an issue with the Python setup on your system, since it sounds like the Python-plotting files are not present. The requirement should basically just be having Python (either 2.7 or 3.*) with Matplotlib ver>=2.2.3. What is the output of this on your computer:

afni_system_check.py -check_all

?
Also, there is a log of the terminal output of creating the APQC HTML in your afni_proc.py-produced results directory, called out.review_html. I will send you a messge through the interface here with my email address, if you wouldn't mind sending me that (which should also suggest/pinpoint what is happening).

--pt