@ LaeMing
In heat engines the implications of the Carnot theorem are unavoidable.
Maximum Efficiency = 1 - T(cold)/T(hot)
http://en.wikipedia.org/wiki/Carnot_cycle
If a secondary heat engine uses the same cold sink as the primary engine had done, then the sink temperature of the primary one will be increased, thus reducing its efficiency. Overall the efficiency is reduced.
Alternatively, if the secondary heat engine uses additional cooling so that the primary heat engine's sink temperature is unaffected after it is coupled in, then it would generally be better to apply the improved cooling directly to the primary engine.
There may be inefficient engines in which some otherwise wasted heat can be turned into useful energy (e.g. internal combustion engine exhaust), but the Carnot limit is impossible to beat; and modern gas turbines already run very close to this limit. This is why Francis KIng was dubious.