Seems very risky. Not something I’d want to do if minimum downtime was the goal. One wrong piece of gravel ends up with catastrophic failure instead of 5 minutes of downtime.
But the goal was zero downtime, not minimum downtime. The client made it clear that 5 minutes of downtime was equivalent to catastrophic failure. So they correctly found a solution that reduced the chance of "5 minutes of downtime", at the expense of an increased risk of catastrophic failure.
