which way monday, page-91

Yes, have checked the numbers in past, some fit very nicely for one of many indices/currencies that can be conveniently choosen from.
Most are weeks or months out though, which for an averaged model is expected.

However, eg...Sep 2000 SPX top into his Nov 2002...find the high at 1 Sep 2000 to SPX low at 10 Oct 2002.
Armstrongs published dates is 13 Sep 2000 and 8 Nov 2002. But then writing 2000.7, gives a whole +-0.05yr range to chose from, ie 36 days, lol, to further fudge the fit.

So not only is there a choice of indices/eco indicator, lets throw in the Baltic as well, there is also a choice of times, up to over an entire month to choose from.

Becoming more and more convinced he already knew a whole bunch of interesting links about π, and whether he knew it or not, generally not if its an obsession, proceeded to search for something out there that contained one of those links.
In his case found, for the cycles he counted, who knows how many might have been conveniently ignored to make his theory 'elegant', just to be able to use the 8.6yrs = 1000π days, which incidently is 3142 days to nearest day, one too many.
There's probably dozens or even hundreds of other attempts he made to find a great π link in nature that were abandoned, or 'explained away' with that great informative rational mind of his, all just so that he could kid himself as being one of the greats.