[RightWhale]

accelleration is always 2-D

Not at all. Spacecraft have control thrusters pointing along 3 axes and they can impart a 3-D acceleration. In the superstring model there can be more than 3 dimensions, maybe even 11 in some cases, and there could be corresponding accelerations.

[MonroeDNA]

I phrased that wrong.

Accelleration can be defined in 1 dimension only, whereas gravity always has more than 1 dimension.

Under accelleration, it doesn't matter if you are two feet higher, two feet to the left, or two feet forward (change in X,Y, or Z location) if you are part of the object being accellerated. The force does not change, no matter how you change your location (static location!)

Under gravity, if you are two feet higher, it changes. If you are two feet to the right (not spherical, XYZ right), the force changes, and can be detected. Under gravity, a static change in location will cause a shift in the gravity vector. That's why the pendulum detection method works.

Gravity and accelleration can be distinguished, under any imaginable scenario. Even in a sealed box, or elevator, IMHO.

I've been asking this question for 15 years, and either I'm wrong (won't be the first time, for sure!), or the elevator or sealed box thing is wrong.

[aruanan]

Spacecraft have control thrusters pointing along 3 axes and they can impart a 3-D acceleration.

But at constant thrust, the acceleration is in one direction as the sum of the vectors.