There is about one half part per million free hydrogen in the air.  It is
reactive, but not quite as reactive as one might believe.

If you have done the old lab demonstration where you electrolyze
water and catch both gases in a bottle, you will see that the reaction
to form water again is neither immediate nor spontaneous under those
conditions.  (A spark or flame is a good initiator, however.)

Ive never analyzed the reaction byproducts afterwards to see how
much hydrogen and oxygen are left, but imagine someone has, and
I would expect it to be higher than 0.5 ppm hydrogen.  Reactions seldom
go quickly to completion.

The inorganic theory of petroleum has been around since Genesis,
I guess.  I have read it, but am not a believer in it as a serious
source of the earth's hydrocarbons.

There is a fair abundance of methane, both bound as hydrates and
free in oil production operations.  It is made fairly easily by microbes
from some carbonaceous sources.  Crystalline clathrates are all too
abundant in some subsea areas, but mining them for the hydrocarbons
is a touchy procedure.  Sudden decomposition of methane hydrate
can lead to disaster.

If you had a magic filter that could separate the atmospheric hydrogen
from the other components with minimal energy, then cheap energy
could be a reality.  If you could separate gold from seawater in the
same way, jewelry and dental work might be a lot cheaper.

There is just no free lunch, is there, Don?