“If you can’t get each job below 4 hours, including travel time, then you can’t complete two in a day.”
A 30 percent speedup looks great in the benchmark and changes nothing in production. The gain has to cross a real threshold to count. Shave a job from 5 hours to 4.5 and a technician still does one a day. Optimization that does not move a discrete boundary or clear the actual bottleneck is just numbers you can brag about in a standup.