I see this too. The people best qualified to maintain stability in institutions are rarely the best people to institute and embed meaningless change - and likewise the people best at instituting change are often not the ones needed to run the reformed organisation after change is embedded.
That's why we see change CEOs established in corporations in various forms - and even there they often fail due to the many unintended and unpredictable consequences of a major change.
I am increasingly of the view that replication is the key to true change - a lesson that a billion years of evolution has taught us. The 'only one of its kind' approach allows no fallback if a particular evolutionary change process results, unintentionally, in a deadend, non-viable solution.
This is the reverse of how most organizations think, but as we see time and time again the markets with more competition are more successful, those that concentrate down to a few players tend to need more external regulation to continue to serve their customers adequately.
For government in particular this is a challenge - it is the ultimate in geographic monopoly, and places barriers on free movement that restrict the ability for citizens to shop around (the highest barriers at national levels).
While our societies may never support the notion of totally free choice in government (which is possible for all but the most geographically restricted services), we can move government to the backroom, an infrastructure platform rather than a service provider.
Then government services can be offered and bundled competitively with private sector services by a variety of providers, creating that competition for outcomes that a single government will never be as good at achieving.