I can't argue with that either, yet it doesn't invalidate its plausibility for me. Call me a cynic. 
Less than 10 years ago, the Georgia Department of Transportation "learned" that the bridge inspections had a really high falsification rate. They had a program where a two-man crew would go out and visually inspect every bridge in the state and rate them on several factors. These would then be used to program capital projects and determine weight posting signs.
The linked article paints this as a human problem of trying to meet an unrealistic deadline that is probably set to performance appraisal. So rather than turn in empty reports they just made stuff up. That said, bridges don't go from a mid-range score out of 100 to a single digit score in one year. There were a couple bridges near downtown that has to be closed and emergency funding programmed post-haste once they finally got inspected.
To me the circumstance in the story is very different than the one you are describing.
I can certainly see how a long-term system with measures put in place to avoid catastrophic failure could be subverted over long periods of time by those participating in the system either neglecting their roles or actively falsifying data because, often, the same government who put the system into place is exerting unreasonable pressure to keep everything within budget where that budget is supposed to include upkeep cost and so from year to year the constant pressure to try to stay within budget pushes back against the necessary maintenance and horrible decisions are made in support of that. I live in Minnesota and witnessed the Interstate 35W bridge collapse firsthand from a boat that would've been passing under that bridge ten minutes later, so I have seen that kind of long-term improper checking of infrastructure first hand and almost been literally smushed by negligence of it.
But that strikes me as very very different than this story, which was a system just being put into place now and was apparently still in the transitionary proof-of-concept period during which I would expect that oversight would never be stricter. The cost of a hardware failsafe would justify its own cost a thousand times over. Even having a secondary human safety level would improve safety immeasurably--most modern metropolitan highway systems already have cameras installed to help automatically monitor traffic levels to aid in GPS traffic routing, so besides the hardware failsafe, and the driver, automatic traffic algorithms could easily sense a conflicting flow of traffic on a two-way road like this and take emergency measures to try to avoid a collision--you could find open source software that could perform a simple utility like this very reliably. It could act by preventing that last barrier from being raised and/or putting messages on configurable signs to warn of imminent collision (these signs are usually available in modern metro systems to say things like "crash in right lane, use caution"). Even if no automatic algorithm is checking this, could pay a DOT employee for a few minutes of time per day to watch the traffic flow for conflicting values just before and after the system is switched over.
Even having two drivers go through the road and call in instead of just one would greatly reduce the chance of one person's mistake causing lots of deaths.
All of these things would be very tiny additional maintenance costs compared to what already is invested to maintain these roads and the cost of cleaning up the mess of a major accident and the inevitable lawsuits that would results would make at least a couple safety measures an important thing to pay, and during this initial just-having-completed-the-project period all of these things are still going to be in the forefront. I could believe them slipping later on when the system has worked so reliably for years and years and years that the budgeters are trying to find places to cut corners and start letting this slip but during the opening trial period I do find that much harder to believe.
I can see why others weren't bothered by this, but this made the climax of the story very very hard for me to believe without criminal negligence on the part of the mother character and everyone involved in the project she works on.
