This is a ψ-epistemic interpretation. It considers all systems as quantum mechanical. In that setting all systems can be needle states or observers. This will then mean a system of observers are on different paths or amplitudes and do not share the same outcome. This is somewhat related to Qbism, where different observers may have different Bayesian regressions or a sequence of measurements. This has some interesting prospects for quantum complexity, where a state |A〉 = H_1H_2...H_n|B〉 for a sequence of Hadamard gates. This defines the complexity C(A,B) as a path or geodesic. The Wigner friend issue means there can be an ambiguity over which path is extremizes the complexity.
This has a weak effect on the macroscopic world. As Heisenberg pointed out there is no sharp boundary between the quantum and classical domains. There is no matter how much we might object some quantum uncertain fuzz around classical reality. This means we could find some irreducible errors in needle states for qubit systems that approach the mesoscopic domain. These irreducible errors may reflects some tiny level of this epistemic uncertainty between observer's account of reality.
LC