The model I am speculating about is more rigidly bound to the intermittency profile of each local wind farm: up to the point that a mining container is not even connected to the grid - but to the farm directly => mining uptime is smaller (depending on how often the given farm is curtailed).
This is a completely different approach than what NFW explained, you are talking about a completely off-grid, isolated mining operation, similar to flare gas mining farms that sit on oil fields. it's a very good model with two big "or rather huge" IFs.
1- How much uptime are we talking?
2- How much does it cost for both aspects, A- power rate B- infrastructure.
For flare gas, uptime is guaranteed until the well is done (usually a decade or more), and you move your generator and container to the next well. it's also super cheap to operate and probably even free because you can sell carbon credits. The generator cost is maybe $1-2M per MW (just a quick Google search so the numbers could be OFF by a lot)
With your setup, we are talking about unstable energy, which then needs some sort of inverter, gearbox, or synchronous generator to get you stable AC power and is not guaranteed to be available all the time.
I would suppose if we are atling about 10-20% down time and the overall power rate is below 5 cents, with infrastructure that doesn't cost a lot more than the traditional mining setup, your idea "may" work.