This antipattern is a flavour of Software project collapse modeled after motor neuron disease:
Unlike the uncontrolled rot of software project motor neuron disease, jenga tower game is a deliberate game of "business optimisation" and "productivity improvement":
The team operates and delivers, thus, pull a piece out of it (delay a salary review, increase the workload, withdraw the monthly cake money): If the team still functions, it reinforces the management's conviction that their approach is correct and even somehow akin to the scientific experiment and lean approach as good as other industry best practices. Pulling a piece out - getting productivity "gain" looks like a cause-effect relation.
Once the jenga tower eventually collapses, not only it is too late. The very last piece removed is seen as the root cause of the collapse.
The impression that pulling out dozens of pieces was success and only the last one a failure further distorts causality.
This is such an insidious and harmful pattern that it is hard to put price on it, unlike other patterns that not necessarily are wrong, but may be just expensive.