Which angle of attack shall we present? The local one sensed by the probe, or a corrected value to fuselage AoA based on flight test with a noseboom, itself subject to corrections. Shall we correct for pitch rate induced alpha, or not?
Then things which you think are "almost constant" often are not. What about variations with altitude, or Mach number? Or cases where scheduled speeds are determined by other criteria - like VMC, for example. Your Vref-alpha chart would be no simpler than the Vref-speed chart.
It's also common to have more redundancy in the airspeed system than in AoA systems - 2 AoA vanes, but 4 airspeed systems, for example.
And to reinforce the previous point - trailing edge flaps (including ailerons) do influence the stall AoA - using a constant fuselage reference for AoA, not accounting for theoretical chordline changes.