This is now my full-time occupation and I've started up a company in its pursuit. 90% efficiency is a stretch but I believe 70% is eminently possible - this is the target of an initial design prototype I am working on. 75% - is the long-term goal and 80% may be technologically within reach. Scientists and engineers generally know the pathways toward achieving this level of efficiency but are currently hindered by cost considerations and lots of ARPA-E and VC money being sucked up by "sexier" technologies than the 150-year-old ICE (forgetting their history that fuel cells, batteries and electric motors are even older tech).
PS: Anyone can play the game of claiming sky-high efficiency figures using freshman 1st law ideal air-standard Otto cycle calculations. I can play the same game too. Every other ICE concept that I have encountered to date claiming efficiencies of 70% or more are thermodynamically implausible and therefore BS.
This statement is true and quite prescient. You might be referring to these researchers:
The practice of declaring thermal efficiency claims from ideal First Law calculations is facile at best and spurious at worst. The researchers in the above propose calculating efficiency from the Second Law of Thermodynamics using the concept of exergy analysis. I tend to agree with this but maintain that we can use both - using one to validate the other. What you might have read suggesting that 90% is possible do so by looking at the sources of combustion exergy destruction. This is denoted "D" in the bottom left of my calculation screenshot above - so yes, 90%
IS at least plausible... However, this will not be achieved burning any conventional fuels in air - which typically have exergy destruction of between about 20-33%, meaning that the highest attainable efficiency is in the order of 67-80% if and only if all other processes in the cycle proceed perfectly reversibly and without any losses due to friction, heat transfer, etc.
