So, if I read this correctly, starting at the top, we have

a dit of five-plus miliseconds of 400Hz (0.00 through 0.05),
followed by an intercharacter break of eight-plus ms of 105.26Hz (audible, but unlikely to be confused with 400Hz),
followed by a dah of something like 18 ms of 400Hz,
followed by a pause of 109 ms at 8.23 Hz (inaudible even to dogs, I think),
... and so on.

The 5+ to 18 is close for the 3:1 dit:dah ratio. the inter-character break of 8 ms doesn't fit so well; and the pause (inter-word break?) is not even close to the 1:7... but maybe it wasn't intended to be. But the next "A" fits less well.

So, if typical CW is like your sample, my theories go up in smoke ...which HAS happened before :-).