I disagree, although I will point out that stainless steel (302, 303, 304, 316) is a specific and (typically) less durable subset of steels (without getting into hardenable stainless such as 440C). The 3x price differential on stainless makes it less than desirable for teams to work with in the raw, aside from the additional tooling needed to properly work it. However, I will support my original contention with the following links and tables...
Reference #1
Reference #2
from reference #2...
........................................Aluminium. ................Steel
Density, ρ kg m−3...............2,700......................7 ,800
Young modulus, E N mm−2....70,000....................210,000
Shear modulus, G N mm−2.....27,000....................81,000
Poisson ratio, ν.....................0.33........................ 0.3
from matweb.com...
1018 steel versus
6061-T6 aluminum versus
ASTM A526 galvanized steel
I could never bring myself to use aluminum for a shaft. It is too weak (comparably) to take that shock and torsion loads, as well as being more prone to unannounced catastrophic failure over steel (steel bends and then breaks ... aluminum has a much smaller deformation period before it snaps under the same loads). I'd rather use a steel shaft so that I could press-fit it into a bearing. The surface properties of aluminum just aren't good enough (at least for me) for rotating or direct-contact sliding.