but I suspect that any attempt to do so without contamination by ambient helium would be exceedingly difficult.
You are wrong. It is not difficult with modern instruments at places like the NRL or the ENEA. Nor is it difficult to capture the helium after dissolving the cathode in acid, which has also been done. Since melting or dissolving shows no significant helium in some samples before the experiment, and significant helium in samples from after tests that produce excess heat, I do not see how you can claim it is "difficult" or unreliable.
The point I am trying to make is that the measured ratios tell us nothing except that helium appears to be a product. Bloggers may see an astonishing coincidence, but maybe scientists would not.
No, the measured ratios show that the helium is close to the expected ratio for D+D => He4, as shown on p. 8 here:
http://lenr-canr.org/acrobat/Hagelsteinnewphysica.pdf
It is even closer when you account for helium left in the metal, which you do by melting or dissolving the metal, as I said.
Bloggers do not "see an astonishing coincidence." The scientists who performed the experiments, Mile, McKubre, Gozzi and others, all claim that the ratio is 23.8 MeV per D+D reaction.