have always thought of Va as the speed where if you applied full control deflection, the surface would stall before it could provide enough force to exceed the max load factor of the aircraft meaning you could safely use full deflection of controls below Va.
Not really, if you fully apply elevator, rudder or ailerons at Va, the horizontal surface, vertical surface and the ailerons must be able to carry the loads - stalling doesn't come into it. See FAR 23.423, 441, 445.
There is no FAR 23 standard which requires the wings to carry the loads generated by full elevator deflection, or flying to the stall at Va.
The exception, in a roundabout way, is when Va=Va min.
In fact FAR25 full elevator application at Va specifically states that "accelerations exceeding the maximum positive manoeuvre load factor need not be considered".
I don't know how the Va v's stall misunderstanding developed, but it goes back many years and was probably contributed to by:
1. Regulatory operations departments and airworthiness not communicating with each other;
2. The standard Vg diagram is really generic and only shows one of the many possible speeds for Va - that happens to be Va min, which of course, is the only stall related Va speed;
3. Operations people looking at the Vg diagram and not reading the written standard which shows Va can really be any speed above Va min;
4. Aircraft manufacturers published Va speeds scheduled against weight for some aircraft. Clearly, control surface stuctural strength doesn't reduced with weight.
But IF the published Va was Va min this approach actually provides some useful stall v's structural strength information. It may be that manufacturers airworthiness staff
didn't communicated wth the operations manual people at this level of detail; and
5. Perhaps many aircraft were designed to Va=Va min, but your example of the Citabria and Decathlon is interesting. Va was called Vp in CAR 3 (circa 1949) and was pretty much the same as the current Va
FAR 25 different wording, but is similar to FAR 23, I don't know what the latest FAA standards will require - it could just require the structure to be "safe" and leave the applicant to convince the authority.I am now also assuming that despite being at or below Va I *can* exceed max G if I use full deflection inputs (is that correct?).
Below Va the structural integrity of the deflected control has been justified, but you could exceed maximum G.
You need to know the basis of the selection of Va. Is it greater than Va min? Also one should consider the various ways you can get a G "overswing" beyond that you expected for the speed you are flying at.