Thirty years ago, I dwelt in the world of lab rats, rat brains, pigeon memory, and animal research in general. Most folks working in that area would tell you that, while there were many similarities between "us" and "them", there were also differences between lab animals and human minds that made it risky to generalize from one to the other. Still, despite the obvious shortcomings of using a Long-Evans hooded rat choosing to turn left rather than right to substitute for a human making a decision based on recollection, we used the rats because we had decades and decades of previous research to draw on and compare to, and that comparison facilitated scientific insight.
All public-sector institutions are subject to accountability to the legislative side, any relevant central accountability agencies, and the electorate. One of the principal tools for that accountability is longitudinal comparison of performance, and longitudinal comparison demands comparability.
Now that is, in no way, a justification for never changing when the proposed change represents an improvement in functionality and service, or value for money. But at the same time, there can be a completely understandable resistance to changing how some things are done if it represents a complete or disruptive break in what can be measured, and how. There is value in sticking with "how things are done" in some areas, primarily because of the comparability it provides, and the insight that can foster. I won't propose that it should trump all, but it's not nothing either.
That is an argument addressing the unanticipated larger impacts of changes that others may be thinking about when one of us has a "great idea". That still leaves a very large arena of changes which do not have such understandable reticence attached to them.