A plane similar to a powered glider, can have a lift/drag ratio of about 40:1, and at 60 mph, have a pounds lifted per horsepower ratio of about 200:1. Now if have two planes flying in a circle, connected by a tether, the two planes can climb or descend in unison. Next, a payload can be connected to the middle of the tether, and the payload can be lifted vertically and landed vertically.
A practical design would likely fly at 120 mph, or faster. The higher speed allows a smaller area wing to be used.
To fly horizontally at a slow speed, the planes could bank in the direction of desired motion as they fly around in a circle.
To fly horizontally at the speed of the planes, first have sufficient altitude, then both planes could go into a vertical dive, one plane could roll 180 degrees about it's tether anchor point. Then, when both planes pull out of the vertical dives, they are both flying in the same direction.
The drag of the payload must be overcome by the planes.
To transition back to circular flight, both planes could go into a vertical dive, then one plane could roll 180 degrees about it's tether anchor point. Then, when both planes pull out of the vertical dives, they are flying in a circle.
The preceding transitions between circular flight and forward flight are an argument that the transitions are possible. The actual maneuvers would likely be less extreme, but more difficult to describe.
For power generation from a thermal, the planes could be diving some into the rising air to avoid gaining altitude. If the thermal is strong enough, the propellers on the planes can generate power.
Most winds blow horizontally. To generate power from horizontal winds, the plane of the circular flight could be nearly vertical. A tether from a ground anchor to the midpoint of the tether that connects the planes would keep the tethered planes from being blown away by the wind. The planes must generate enough lift to stay airborne, as they would from a less than vertical plane.
The above mirrors a windmill. The planes are like the outer portions of windmill rotor blades. The tethers connecting the planes are like the inner portions of windmill blades. The tether to ground is like a windmill tower.
The tip speed of a typical windmill rotor blade is ~150mph. The planes would fly at a similar speed. A rotor (prop) on each plane could be much smaller than a windmill rotor, and need much less gearup for power generation. The first plan for power generation was to use a generator in the plane, powered by the prop.
The second plan for power generation is to provide a generator of enough size in kilowatts to absorb the prop power from a moderate wind. Then for a stronger wind, the additional power from the prop could drive a compressor to compress air to about 60 atmospheres pressure. Then pipe the compressed air through hollow tethers to ground. The compressed air could be used directly on the ground for electric power generation, or stored in underground caverns for use as needed. This would store compressed air for when the winds are not blowing. Note that considerable heat energy is lost in compressing the air.
For a plane speed of 150mph, 220'/sec., I calculated a 151.25' radius of the flight circle for a 10 g inward acceleration.
On the ground, heat could be added as the air is expanded to generate power. Solar heat is most available when the demand for electricity is highest.
Overall, I surmise that the overall economics are in favor of the combined generator and compressed air approach. My guess is that this mode of generating power from wind is likely a few times cheaper than current windmills. The cost of the tethered planes should be a few times cheaper, per kilowatt, than a conventional wind turbine.
The compressors on the planes could act as motors. This could keep the planes airborne for short times of not much wind. Also, for VTOL of the planes, something like the design of the "Goldeneye UAV" could be considered.
For more information, google on "tethered planes Peter Lynn" without the quotes.