I hate to do this, but it's time for Thermodynamics 101. According to the second law of thermodynamics, there is an upper limit on the amount of work that any device can extract when operating between a high temperature source and a low temperature sink. This is called the Carnot Efficiency and can be calculated as follows:

Eff = 1 - TL/TH

where TL is the absolute temp of the low temp sink and TH is the absolute temp of the high temp source.

Accomplishing this efficiency would require a "reversible heat engine" which is a theoretical device that is completely reversible (no heat loss, no friction, ideal transfer of thermal energy to mechanical energy). No practical device approaches this limit, certainly not an internal combustion engine.

If you put numbers in this equation (and assume about 70F ambient temperature) you will find that 70% efficiency can be achieved with a high temperature source of about 1300F. Therefore, it is possible to have a 70% efficient engine without violating the second law of thermodynamics. So I will not tell you it's impossible.

However, the "engine" required to accomplish this does not exist. The internal combustion engine can be approximated by something called the "Air-Standard Otto Cycle" which can be thought of as an ideal internal combustion engine. The efficiency of this cycle is significantly lower than the Carnot Cycle due to the fundamental design of the engine. In addition, real engines have significant additional losses (i.e., every BTU of energy removed by the cooling system). The current state of the art for actual internal combustion engines is around 30%. Much of this limitation has to do with the nature of the Air-Standard Otto Cycle. Another significant limitation is the maximum temperature that current materials can withstand. Obtaining a significant improvement will require a change to the fundamental design of the engine. This has lead to proposals such as ceramic engines with no cooling. By way of comparison, a state of the art power generation stationary combustion turbine can achieve efficiencies in the range of 50%. This requires operation at the extreme limits of the materials and some fairly exotic designs. We are talking about installations that will spend (literally) millions to increase efficiency by a couple of percent.

The bottom line is that no-one on the planet has an actual practical device (engine) of any type that will convert thermal energy (at these temperatures) to mechanical energy with anything close to 70% efficiency. Very little of this limitation is due to the combustion process, which is the only part of the process that the mythical "100 mph carburetor" could improve.

Sorry for the rant, but it annoys me when I hear statements that directly contradict undergraduate physics presented as fact. This nonsense has been around forever, and I don't think we need to propagate it any further.