[citation][nom]Abrahm[/nom]Actually, that's not necessarily the case. Plants generally only have an energy conversion efficiency of around 3-6%.[/citation]
You are correct for conversion to biomass (i.e. plant growth). That number also takes into account metabolic usage by the plant. Chlorophyll itself converts about 30% of the collected light into usable energy, the glucose building and consumption process' inefficiencies and plant metabolic needs reduce it to 3-6% net gain. It gets even worse when you try to convert that biomass into energy (e.g. bio-diesel), it ends up reducing it to around 0.3% efficiency. If we can remove the 'plant' part (so to speak) from the bio-fuel process it could go a long way towards improving the process.
By comparison, current semiconductor based transistors have lab efficiencies of 45% or so with multiple junction units under ideal and controlled conditions. However, once you put them in the field, they tend to drop to 8-12% (heating being one of the problems), and that's only for conversion to electricity. Trying to convert that to transportable power (bio-fuel or battery) reduces the efficiency even further.