efficiency is always important.
Your available usable power is the product of efficiency and non usable (without tranformation) available power.
zero times anything is still zero
10% times anything only allows 10% of the energy use...
But I do understand why this is being framed in this way. If you are going to convert these types of power sources you cannot expect to get to conventional efficiencies considered norms in normal power suppl design.
Thus the lowered expectations framing of the issue.
The numbers in the Table for solar panel don't look right to me. With solar constant at ~1400W/m^2 and accounting for atmospheric attenuation, albedo, etc., let us say an insolation of 800W/m^2 which works out to 0.08W/cm^2. With solar panel efficiencies at 10 to 20%, this is few 10's of mW/cm^2, NOT few 100's claimed in the table.
The article otherwise a smorgasboard of Linear's solutions!