I have reworked this using accurate data. It looks like the difference was mostly due to me making a mistake with the nav computer, rather than differing assumptions. CAS has to be almost the same as EAS, even the online calculators show only a knot and a half of compressibility, which has no real effect. I have reworked it several times to find M0.420 so have amended the answers above in red and annotated previous posts to point out my mistake.
Interestingly the CRP5 nav computer is still giving a TAS some 4KT higher than the online sites. I remember this discrepancy was noticed some ten or so years ago when the JAA exams first started, other nav computers gave different speeds. It led to the UK CAA at least saying we should use the Pooleys or Airtour CRP5 for TAS calculations as that was the one they used when they wrote the questions.