That is a good point, but would be less of an issue now, with the new set of V5 sensors coming soon: Coming Soon - VEX Robotics
That may depend a lot on the environment. Here is a good example: Sacramento State Management & Venue Problems
When VEX makes sensors they don’t have to guarantee that all of them will work 100% reliably in all possible circumstances. That would be way too expensive. Those sensors just have to be good enough not to randomly fail and be similar in their limitations to the cheap counterparts that are widely used in the consumer grade electronics.
The true objective would be to give students a choice of sensors that each have their own constraints and limitations and then have students go through the design process to select the ones best suited for their task.
Now that 3-wire 8 port constraint is no longer an issue, I would encourage teams to add all: ultrasonic, Infrared, and mechanical limit switch sensors to their robots and then test and document the pros and cons of each.
That would look amazing in your EN and, for extra credit, you may program your robot to look at the observations from all sensor types, and determine at the run time if any of them return suspect values and ignore those.
I would bet on limit switch being the most reliable, but you never know.