At any given time the atmosphere as a whole contains about 0.25%, by mass, water vapor. With an atmospheric mass of about 5x10^18 kg that puts 1.3x10^16 kg of H2O in the air. Water's latent heat is 2.27 MJ/kg, for a total energy content of close to 3x10^22 J, or 75 nominal years of present-day energy use. Since the water vapor energy used in a year is 3000 times human energy use, that gives us an average stay time for water vapor in the atmosphere of close to 9 days, over which time all the water condenses out and is replaced.
However, the bulk (about 2/3) of Earths fresh water is stored in massive ice sheets which themselves hold considerable gravitational potential energy due to their height above sea level. The total ice sheet mass is about 2.4x10^19 kg. With an average height on the order of 1 km above sea level that amounts to 2.4x10^23 J of gravitational potential energy, or 600 years worth of present energy use. One can imagine new hydroelectric technology to harness this vast store of energy as the ice melts in a warmer world.
The salinity difference between fresh and salt water is another potential source of energy - it is claimed that 1 MW can be derived from the osmotic pressure associated with bringing 1 m^3/sec of fresh water in contact with ocean water. That amounts to 1 kJ per kg of fresh water; the full store of fresh water on Earth, 3.4x10^19 kg, then yields 3.4x10^22 J or about 80 years of present energy demand just from the salinity differences between fresh and ocean water. Two thirds of this store of potential water energy is again in Earths ice sheets; if we just limit ourselves to fresh liquid water on the surface (lakes and rivers), only about 1 years worth of energy is on store at any given time.
Whether through expansion of traditional gravitational hydroelectric power, through some mechanism for harnessing the latent heat of water vapor, through salinity differences or some other mechanism, Earth's hydrologic cycle has enormous potential to power the world. It may face great technical challenges in meeting that potential, but the physical scale of the resource is very large.
Winds are powered from solar energy both directly through the differential heating of air, and indirectly with the release of water vapors latent heat as it condenses to form rain droplets in the atmosphere. If Earth's entire atmosphere moved at the low average wind speed of about 20 km/h (5.6 m/sec), the total kinetic energy, 1/2 m v^2 would amount to 8x10^19 J, or around one fifth of a year's worth of current energy demand.
On average winds around the planet dissipate power at a rate of at least 200 TW. Several studies(1) have shown several tens of TW could be practically extracted over or near land with near-current generations of wind turbines. So wind represents a resource a few times the present scale of world energy use - sufficient to power the world for a few decades at least.
(1) Cristina L. Archer and Mark Z. Jacobson, Evaluation of global wind power, J. Geophys. Res., 110, D12110 (2005), which estimates a global land wind resource of 72 TW at 80 m altitude.