At a very rough estimate, what proportion of wildfires are caused by natural events like lightning? I am absolutely sure that in Europe, where there are often wildfires, the sun is never hot enough to set fire to vegetation directly, though it obviously makes things tinder dry and easily combustible. The broken bottle theory has been disproved. Experiments involving broken glass among highly combustible straw or hay were unsuccessful in causing combustion, though manual focussing of the suns rays with the lens-shaped bottom of a bottle were on rare occasions and with much difficulty successful. Although lightning undoubtedly causes the occasional wildfire, the majority have no connection with lightning. It should also be remembered that lightning is usually accompanied by torrential rain. These fires often start in several places at once, which is stretching the long arm of coincidence a bit too far.
Apart from the very rare instances of volcanos causing wildfires, what other natural causes could there be? Comparison of wildfire frequency in tinder-dry uninhabited areas, well away from civilisation, versus frequency in or near similar but populated areas should give a good idea of the scale of the problem.