Detecting (absolute) acceleration in an almost empty Newtonian world

Newton believed that there was an absolute spatial frame of reference:

Absolute space, in its own nature, without regard to anything external, remains always similar and immovable.

So as far as Newton was concerned, acceleration could always be detected relative to this absolute reference frame, because Newton's laws of motion would not hold in an accelerated reference frame (unless fictitious forces were introduced). So even in a universe with only two objects, it would always be clear which of them was being accelerated.

On the other hand, Newton's contemporary Leibniz believed that there was no absolute space, and that space only made sense as the relative location of bodies. So for Leibniz, Newton's laws of motion could only apply in a universe in which there were enough distant objects more or less stationary with respect to each other (the "fixed stars") to define an inertial frame of reference. In a two object universe Newton's laws of motion would not apply. If there was relative acceleration, both objects would see themselves as stationary and the other object as accelerating.

Ernst Mach expanded on Leibniz's point of view and encapsulated it in Mach's principle. He is reported to have said "When the subway jerks, it's the fixed stars that throw you down". Mach's thinking was a guiding factor in Einstein's development of general relativity.

How about if we say this: A powered change in acceleration can be experienced and measured. An acceleration caused by gravity can only be measured relative to the associated sources of gravity. Which is to say, If you were in an enclosed lab in a space ship in free fall (with the engines off) there is no way to determine the acceleration you might be experiencing due to the gravity of a nearby planet or star. (When I refer to a “change in acceleration”, I mean that it is in addition to the continuous accelerations that we are subject to from various concentrations of mass.)