A tacking sailing craft, whether it be a boat, sand yacht, or ice yacht has to react a force vector from the sails against one from the keel / wheels / skates to provide a component of force in the direction of travel.
If the direction of travel is directly into wind, this is not possible. The force vector from the sail can't act at better than right angles to the wind. And even if it could act at right angles (which requires an infinite lift/drag ratio) it still wouldn't create any forward component when combined with the reaction force from the wheels (which again can't act at better than right angles to the direction of travel, even given perfect frictionless wheels and zero energy loss grippy tyres).
For the same reason, you can't just use two or more sails reacting against each other. The sails each provide a sidways 'lift' component, and a backwards 'drag' component, and there is no way to combine these to end up with the required forward facing component.
I suppose you could have one or more tacking small wheeled vehicles pulling the main vehicle. This would be the wind powered equivalent of a horse drawn waggon, except that the horses would be running zig-zag.
Another method would be to fly controllable kites (two or more strings) from the vehicle, and use the varying line tension, and angles to power the wheels. This is really just the same as the windmill, except that you replace the windmill with kites. I don't think it would be anywhere near as practical as the windmill method.
I like the idea of a vertical axis windmill (like a cup anemometer, or a Savonious rotor). This would provide forward drive irrespective of wind direction - with a horizontal axis mill, it would be necessary to point the rotor into the direction of the local airflow, requiring complex machinery - the vertical axis method eliminates this, though at the cost of efficiency.
