The “infinite source” sounds rather grim.  Clearly the Himalayas will continue to provide sediment to Bangladesh for a very long time.  However, the principle point of the paper is that arsenic is likely mobilized from sediment at the near surface, not deep in the aquifer where wells are extracting groundwater.  The geochemistry of the water and sediments from the aquifer indicate that oxide reduction is no longer occurring in the aquifer.

The basic finding is that arsenic concentrations are not in steady state.  Because  pumping increases flow through the aquifer, and because recharge is free of arsenic (except for irrigation return), one might expect arsenic concentrations to decrease.  This is the idea behind pump-and-treat in the U.S. and Europe, but there is no treat in Bangladesh!  However, the increased organic load from ponds and rice fields may also mobilize arsenic at greater rates.