The elephant in the room with fanciful dreams of high efficiency high compression SI engines is that they cannot run on low octane fuels without fuel preignition. In reality all SI engine compression ratio's are limited by the detonation margins of the fuel in question.
And before folks weigh in talking about how great their 10.5:1 engines are I expect
none of these higher compression NA engines have been tested to
AC 33.47-1 - Detonation Testing in Reciprocating Aircraft Engines (faa.gov).
These tests
require that the engine be operated to within 10F of its maximum CHT and +20F hot day conditions. To pass the tests engines must then be set to operate at least 12% richer than the detectible detonation mixture at full power at max temps. As far as I know, no certified engines have passed this test on 100LL fuel at >9:1 CR, let alone E85. Dreaming of 14:1 CR and higher SI engines in planes is just that. Dreaming.
The AC 33.47-1 tests are performed at sea level however if the engine is turbo or supercharged then engines must also be tested at all operating alts. This results in turbo engines having
even lower CR ratios than naturally inducted engines and so automatically their efficiency is even worse.
By way of comparison, Diesel engines are unaffected by these detonation limitations and can be tested to optimum stoichiometry without the need for the 12% fuel margin. They are also unaffected by turbocharging and can maintain high compression ratio's with subsequent high efficiency whether turbocharged or not. Their mission fuel consumption is therefore much lower as they can maintain full or partial power at Lambdas of 1 or less. For the CD155 Technify this equals around 0.36 lb/hp/hr. By comparison
TCM's own SID-97-3E shows that the Continental TSIO550G has a BSFC at 0.65 lb/hp/hr at rated power or 82% higher than the Diesel. The TSIO550 does improve when leaned for cruise however the overall mission consumption is still hurt by the rated power fuel requirements which include the mandated detonation margins.
Expecting that partial power automotive peak stochiometry data can translate to real world aviation ignores the issues that come with continuous high power operation and the mandated detonation margins required therein. If experimental folks don't test to AC 33.47-1 in order to get better BSFC then they must accept the risks of running their engines that way. It follows that statements like "Otto/Miller/Atkinson is more efficient than Diesel" ignore the realities of AC 33.47-1.