>>9709468At a given wavelength, angular resolution is inversely proportional to the size of the mirror.
James Webb 2.4 meters
Aricibo 305 meters
So optical Aricibo would show 127 times the detail.
You've seen that photo of the Earth taken by one of the Voyagers looking back from the outskirts of the Solar System? We're just a single pixel. Pale Blue Dot. If the Webb took a comparable picture, then OA's image would be 127 pixels wide; good enough to make out continents and weather (assuming we could see the day-side of the exoplanet.) And the exposure time would be correspondingly shorter.
The flip side is that mirrors must be shaped to a fraction of a wave length of the light they're working with. Since radio waves are many times longer than those of visible light, Aricebo's surface curvature is relatively crude, but it still works. Shaping a visible light mirror 300 meters across is probably infeasible at present. Not to mention the cost of lifting it piecemeal into orbit and assembling it there.
