http://www.ncbi.nlm.nih.gov/pubmed/12471319 http://www.ncbi.nlm.nih.gov/pubmed/12471319 quotes gross mechanical efficiency at 24.5 +/- 0.7% for world class cycl

http://www.ncbi.nlm.nih.gov/pubmed/12471319 http://www.ncbi.nlm.nih.gov/pubmed/12471319 quotes gross mechanical efficiency at 24.5 +/- 0.7% for world class cyclists. More importantly, however, http://www.ncbi.nlm.nih.gov/pubmed/15241718 http://www.ncbi.nlm.nih.gov/pubmed/15241718 finds no differences in efficiency between world-class and recreational cyclists. In any case, the "cycling" bit is somewhat of a smoke-screen; it just so happens that it's most easily measured and objective data has long been available through power meters. There is no free-lunch, regardless of the particulars of any sport. (I should however probably clarify that the power numbers quoted above are mechanical power as measured during cycling, which is pretty much the only sport that even has ways of measuring power reliably)