Final answer:
Once utilization exceeds 90%, the waiting time becomes more sensitive to any further increases in utilization because the system has less spare capacity to absorb new work without significant delays.
Step-by-step explanation:
Once utilization is greater than 90%, the waiting time is more sensitive to increases in utilization. This is because as utilization approaches 100%, the system has less capacity to handle additional work without increasing delays. When utilization is already high, any additional load on the system (or increase in utilization) can cause a disproportionately large increase in waiting times. For example, suppose a server can handle 100 units of work and it's currently at 90 units, which is 90% utilization. Adding just 10 more units of work brings it to 100% utilization, potentially causing a significant increase in the waiting time for those last units of work.