Could be. As Jason points out, the engineering use of "accelerated" flight is a misnomer. Acceleration, in the pure mathematics of physics, is nothing more exotic than the first-order time derivative of the velocity vector in some reference frame. It can be computed just as easily for pure orbital motion as it can be for powered flight. So from a pure physics standpoint some of those thought experiments still make sense. If the reference frame is, say, some part of the ISS, then some object moving within it and reckoned according to it could exhibit certain "pure" Newtonian motion -- if the measurement is casual and air is ignored -- even though the whole kit and kaboodle is whipping around the Earth in orbital motion. That's why any question regarding velocity and acceleration is met with the retort question, "...reckoned according to what?"
For powered flight very near some large astronomical body, the body is often considered the fixed point against which velocities (and their accompanying accelerations) are measured, because orbital motion around that body is what dominates your dynamics and engine maneuvers are most often designed to affect that orbit. You ignore the effects of other, more distant, orbiting bodies, or bodies around which your entire system is further orbiting. For orbits around the Earth, Earth can be considered fixed. Same for the Moon. For transfer orbits, you need a multi-body solution. In Apollo this was too much heavy lifting for the AGC, so it used an approximation of what's called the "restricted" three-body problem. In that model, the two principal bodies (Earth and Moon) together determine the orbital path of the third body (the spacecraft), and the spacecraft's mass is assumed to be negligible -- i.e., the paths of Earth and Moon are not measurably affected by the third body. In practice this was implemented first as an Earth-fixed model with adjustments for the increasing effects of the Moon's gravity, then as a Moon-fixed model with adjustments made for the decreasing effects of Earth's gravity. There comes a point at which the computer switches models and, due to the errors in approximation, seems to "jump" instantly from one reckoned position to another. Then of course you have observations from Earth of the spacecraft's position and velocity along the vector from the receiving antenna, so that provides a toehold for Mission Control mainframes to more finely fix the spacecraft's state vector, which can then be "poked" as needed into the AGC.
It's important in understanding Apollo engineering to realize that the computer had to use these various synthetic models deduced from orbital mechanics, and then separately a a closed-loop measured method when under power. The orbital model needed to update the state vector only once every several seconds in order to maintain an accurate knowledge of where it was in space. But under "accelerated" flight, you needed to poll the accelerometer data ten times a second or so in order to capture any vicissitudes of engine operation and update the state vector. They're qualitatively and quantitatively very different solutions -- both necessary in order to fly in space. But it masks the purity of what we sometimes consider in thought experiments carried out in space.
