I will take your word on that
A chi-square test of independence showed that there was no significant difference between the failure rate reported on B6G between the 2.7 and the 2.3 engines (2.7 or 2.3), X2 (1, N = 12) = 2.4604 , p > .116752 .
I'm using your numbers of 12 failed engines and possibly 20k produced. The number of 2.7 orders is approximately 73.6%. So I used that number for the sample size for each engine. For the 2.3 I had to use 1 as the failed engine variable because you can't enter zero and get anything useful. Using that extremely and terribly unreliable data, I ran a chi test. That showed it is not significant. If there were 30,000 produced, the p value would be even less significant.
This is all hypothetical and made up data so it is completely unreliable and meaningless other than the fact that the difference is so tiny it isn't in .05's neighborhood.