In classical physics and special relativity, an inertial frame of reference is a frame of reference that is not undergoing acceleration. In an inertial frame of reference, a physical object with zero net force acting on it moves with a constant velocity (which might be zero)—or, equivalently, it is a frame of reference in which Newton's first law of motion holds
Italics mine.
The moon around the earth is not in an inertial frame, as it is undergoing acceleration, $d\vec p/dt$, $\vec p$ the vector of momentum, going around the earth. Thus even in the Bohr model which has the electron rotating about the nucleus, the electron is not in an inertial frame.
Edit after long discussions in comments:
Arc in a comment:
theories overlap, so we should make them at least partially consistent, that's why I believe we should be able to speak of inertial frames in quantum mechanics, at least in some sense that matches classical mechanics on these boundaries
Quantum mechanics is a probabilistic theory. It can only calculate the probability of a specific event happening at $(t,x,y,z)$ , with the four vector at $(E,p_x,p_y,p_z)$ for energy momentum through the calculations of the wavefunction of the system.
The classical and quantum frames overlap for free electrons, whose "track" can be fitted with classical formulas in the detection chambers. An electron with no fields in the chamber goes in a straight line, and an inertial frame can be calculated. ( putting accelerometers is out of the question because quantum mechanics enters and the state of the electron changes by the existence of a macroscopic machine with billions of molecules.)
The theories should overlap and give the same predictions where they hold , within the limits of the variables where they are valid. Quantum mechanics theories became necessary because Newtonian mechanics was unable to describe the measurements in small dimensions, at the level of single electrons , molecules etc.
Quantum mechanics,was invented as a theory eventually , starting with the inability of classical mechanics and electrodynamics to model black body radiation, the photoelectric effect and atomic spectra.
A bound electron, if measured, at time t will have a probability of being found at rest with the laboratory, at a given (t,x,y,z) that is all. Probability. where it will be at t+dt is also a probability. That is why there are no orbits in the atom, but orbitals.

= <F(x,t)>$, so for highly localized momentum it approximately follows the classical acceleration, but with caveats. For a visualization see these wavy plots.
– Arc Feb 19 '22 at 16:58