yes, it is not a true mathematical law. on the other hand, i have loads of circumstantial evidence (in the feedback and suggestion forum as well as gameplay discussion) that prove my point, in a roundabout way.
"intents and purposes" is not some shocking new turn of phrase, having been used at least as far back as the 1500's see: http://en.wiktionary.org/wiki/for_all_intents_and_purposes
feel free to combat it, it is not for me to stop you from your desired futility.
I might be wrong, but this post gave me the impression you missed what /dev/humancontroller was saying: that an unreserved direct proportion would mean that eventually the event would be
more than certain and that this makes no sense. This interpretation is pretty pedantic though, since by the same token you can infer that the proportion must be bounded by probability = 1 and that this was left out for brevity.
Such a proportionality can exist, for example in a game of six rounds (numbered 0 to 5) where at each round you roll a die and if you roll a number larger than the number of the round you progress to the next round (progressing passed round 5 is a win). In this game the probability of failing to progress at any given round is directly proportional to the round's number. If you like, you could even add a sixth round so that nobody could win

. Note that this is not the same as saying the probability of
reaching a round is proportional to the round's number, since this event depends on having reached all the previous rounds as well.
Now, this isn't to say that this game bears any resemblance to a message board. Far from it. Since the content of posts depend on so many real-world factors, you're unlikely to find such a simple relationship. That is why it would be natural to assume you're simply
modelling the relationship, which allows a measurable error for the sake of simplicity. In this case you would need to better define the event of '[having] degrade[d] to pure flaming', but I would guess there exists a dependence between the occurrence of a flame-like post and the subsequent number of flame-like posts (without considering ragequits and thread locking) so your model might be somewhat justifiable. It should be appreciated, though, that even without such a dependence the probability of
almost any event occurring increases over time by the Infinite monkey theorem.
Yet I'm sure that your actual intent was even more basic than this: You were making an analogy.
As time goes on the probability increases, just like two variables in direct proportion. But a mathematical relationship is a bad choice for an analogy: the analogy fails to illustrate your loose meaning. Maths is a formal construct and by referring to it you can only be asking for a literal interpretation.
As for "intents and purposes" I don't think he thought this was a new phrase. He was just complaining about how people use it. I'm just glad you didn't say "intensive purposes". For me, though, this is way down the list:
I'll teach you to annoy me, because I could care less about your choice to literally sound like a broken record in the way you parrot phrases irregardless of whether you understand them.