> No, they're 30% of nameplate capacity
To translate that to the more wilfully clueless, that means that in REAL PRACTICAL terms their annual power output is 30% (or less) of the "nameplate" capacity (ie: that 1MW wind turbine puts out an average of 300kW when averaged over a year - meaning you need 3 times as many as you thought - when you thought you needed 800 to match ONE gas-fired (or nuclear) power plant you really needed 2400 and in fact you needed 2400 in one location and 2400 in 2-3 other locations, plus battery backup systems to ensure stabilised power output, or you're pushing the costs of coping with wildly intermittent power sources onto the distributors.
Once intermittent sources exceed 20% of the grid, things get hairy - that's what caused the South Australian power blackouts. You can assume that at that point, subsidies are going to be pulled and grid operators will _insist_ that renewables suppliers have battery stabilisation systems on _their_ side of the connection.
> and require spinning conventional backups for when they're not available"
And if you're not willing to pay for those spinning backups to be maintained AND pay their minimum startup costs, you get wide area blackouts - it's exactly that reason which was responsible for the SA blacklouts (the forecast drop in wind wasn't long enough to allow the generated electricity to pay for the backup plant startups, so the owners declined to fire them up and lose $5million or so. Lights went out until payments were guaranteed.)
Renewables are hopelessly expensive when you put all the hidden costs together - the amount of subsidisation that National Grid is forced to provide via "must take" rules alone is higher than the direct subsidies. It's no wonder that windfarm operators are being paid £30k/month per turbine to NOT connect them to the grid (ie, what's being farmed is subsidies)