This is unfortunately the reality of human behavior. The more defined a probabilistic forecast is, the more expectation there is of it. The NWS can say a million times that the center of the cone is not a track prediction, but people by and large cannot interpret probabilistic risk.
We've never had things better in terms of lead time and prep, but as the forecast verifies more and more often, the more people will be angry about vacillations in increasingly less significant details of that forecast. Less significant in terms of public policy and risk management, not less significant in terms of their house.