Bearcat,
In Section E.2.1, Section C.3 is referred to, which indicates that all members that participate in the stability of a structure shall have an effective length factor of 1.0. I would take this to mean that cantilevers do not fall into this category, and fixed-end columns are to be conservatively considered as pinned, thus k = 1.0 (unless you can justify its fixity). This is all that I can ascertain as to the reason why higher values of "k" are not used. I am not sure why it was changed from the 2005 version to the current 2015. 2010 is also like 2015 in this regard.
Buggar,
I have found that the Aluminum Association is both approachable and responsive, if by responsive you mean willing to back up what they publish with technical answers. Changes to their code occur (review, revisions) every 5 years and all changes pass through an Engineering and Design Task Force (EDTF), consisting of professionals in the field of aluminum structural engineering. The proposed changes, if passed by the EDTF, are sent to an Engineering Advisory Committee, consisting of more structural engineers who have experience in aluminum and who make up a broad range of interest in aluminum design. Only after a change is approved through both committees can it become part of the code.
On page 1-4 (2015), there is contact information to send technical inquiries about the code, and I feel certain that you would receive a response if you polled them about the change we have discussed here. Thanks--