There's one problem with the numbers given here:
I have recieved non-triple-spins in GHQ, and quite a lot of them. That means they're not triple-spin-guaranteed (Either that, or I managed to hit a bug).

Also, the last 800sta I spent today in BA has netted me:

• All three chance missions (18 chance times)
• Pero and Exp one-star missions (3 triple-spins each)
• Around 10 to 15 triple-spins that didn't count to current missions (Pero and Exp)

That's almost 40 triple-spins out of 800sta, or which is arount 10 times more triple-spins than you are accounting for.
Even considering today was pretty much a nce in a lifetime level of lucky spins for me (I do admit these numbers are almost unveliebable for BA), luck drifts could never account for a 1000% difference between measures.

That means that your values are just as skewed (to the bad luck side) as mine (to the good luck side).

Once again, that's why I say that for these values to be relevant we'd need to conduct a real long experiment, with more than 2 accounts, and definitely more than 12hs of stamina (or less).
Just to make a point on this, I can tell you that for the whole Guards event I drew 20 to 50 N-Gacha tickets every day, and only saw Akie thrice (and, of course, didn't get her). However, that doesn't mean the probability of getting her was above 1:600, seeing as a few thousand people got her in the first 100 N-Gachas. Thus, their 1000 results of 1:100 have a lot more weight than my 1:600+.

Another example: There's people that got a limited card from Kurito on their first 5 draws (I got Aoe in the first one, and I only gave her a Peranim or Peropudding, can't remember), while there's other people that have been feeding her daily R-Gachas for months and haven't even seen one.