Not to belabor the obvious, but isn't that the purpose behind these discussions? If we don't recognize differences and take pains to work around them all we end up doing is talking past one another.
I haven't noticed a definition of 'knowledge' put forward and agreed upon in any of the earlier posts. You separate the terms 'knowledge' and 'certainty' in your original post but are they not inextricably linked? If we simply accept any input into our consciousness without first calculating its inherent level of certainty we can not call it knowledge as opposed to say, 'rumour'.
We can categorize any number of informational inputs with arbitrarily chosen divisions without statistical limits imposed but if we do are they relevant? Should we not rather first set the statistical lower boundary of certainty in the definition of knowledge and then instead of trying to specify each category's level of certainty, specify each category's relevance to our world view?
Unless of course we are just trying to compose a list of informational inputs.
And yes I know, I tend to babble. But it makes me endearing, just like a puppy.