Assume that some of their models could, for some types of mines, in some deployments and environments.
But that is very different from saying "a Boston Dynamics robot dog is a reasonable primary tool to use in defusing mines".
>Robot dogs are taking over the US military
>“What you’re seeing is a growing array of sensors and effectors as these robot dogs take on more and more roles,” Singer said. “The legs are agnostic to what they’re carrying, whether it’s a soldier’s backpack, a chemical weapons sensor, or a .50 caliber machine gun.”
"Could" they?
Sure - some could
But they're probably not - overall - the best tool for the job (the mine-clearing flail that mounts to the front of an M1 Abrams is far more cost (and time) effective (at least on major routes)
A robot's probably better than a well-trained soldier in most situations - if a robot gets blown up, it's only a dollar figure, not a dollar figure, time to train, grief among the unit, informing family and friends, etc
The question will really come down to "why 'defuse' vs 'detonate'?"
And then, once the "why" has been answered, are the follow-ons - how soon, how "accurately", how many, etc