Estimations of the potential generation resource of wind power happens at all scales from very coarse global estimates, to detailed estimates at individual sites based on specific locations and specific choices of turbine. And this happens at different timescales, from seconds to decades; and both onshore and offshore. It happens in real-time for forecasting minutes, hours, days ahead; and it can happen for farms not yet developed, as part of the design process.
What part of that process of estimating wind resource has the potential greatest benefit from further research?