This is what I mean by knowledge, if you can come up with a better way of explaining it in terms of this site where there is no certainty other then what I have explained then please do so.

You seem to be meaning two things by knowledge, depending on the context in which you use it. I would suggest that you might find it easier if you use the words 'information' or 'evidence' when talking about justifications for a level of confidence/faith. And only use 'knowledge' to signify whatever high degree of confidence you've decided to use as your cut-off point for hands-in-fire-get-burned, I'm-looking-at-the-computer-and-it's-still-there levels of certainty.

It still seems to me that you're going to end up with problems if you hold faith as being another word for confidence. Since even certainty (100% confidence) is still a degree of confidence – and also knowledge under such a definition. But scripture holds that knowledge isn't faith – which is the same as saying if you hold that faith and confidence are synonymous – that knowledge isn't a degree of confidence (even 100% confidence).

It seems to me to be a deeper problem than one of definitions.

You're going to have problems if you say that faith just refers to the preceding 99.9 recurring % levels of confidence, too. You've said that it's a general enough idea, in “Lectures on Faith”, to just be taken as a synonymous term with confidence. But even putting that problem aside, when you wanted to start talking about scripture again I suspect you'd end up saying either, 'People don't believe that which they know.' Or, 'People don't believe that which they have faith in.'

But belief isn't one of those fuzzy terms, like knowledge or faith. The meaning can't be altered to fit a particular argument without doing significant damage to the network of references into which it fits. If I say I believe my computer is in front of me while I'm typing on it, (which going by your standards would also be knowledge,) then there's no significant question what I mean. Just as it's coherent for me to say that I believe my front door is locked, when strictly speaking I've heard one of the other occupants of the building come in and do not know if they locked the door. You might ask how strongly I believe it, but it's coherent for me to answer that in degrees of doubt/confidence; even to the point of saying that I have no doubt.

If you start saying that belief means something else to patch the problem in the epistemology, then you're going to have to explain how that new definition is coherent with all the other instances of its use. And that will alter more, even more well defined, meanings to do that which you're going to have to redefine along with everything that hooks onto those instances... and so on until eventually you've defined everything in terms of whether it makes your theology right. It doesn't seem to be a workable approach.

what do you mean by meaningful faith?

Better odds than the chance a similarly but non-testimony privileged observer has. Or to put it another way whether the faith adds any information. (With non-adding faith being meaningless.)

Take the coin-flip example:

If we were to say that somehow divine testimony could predict the outcome of a coin flip. (Not that I'm saying it can but if it could - or someone claimed it could.) You might get three groups – in separate rooms – and one group would commune with god and score down their predictions, and the other two groups would flip coins. And you'd see what the difference between the god group and one of the coin flipping groups was as compared to the difference between the two coin flipping groups. Do it a few hundred times to get the errors down to whatever you'd decided the noise level was and see whether the faith group was more reliable.

If at the end of all that you didn't have a higher degree of confidence in the predictions of divinity than those of a coin flip, then – assuming that was all the evidence you had for god – you may as well use the coin flip to dictate your actions. The faith wouldn't add any information, it wouldn't hook onto the world you were experiencing.

You could do the same sort of thing with the more day to day predictions of prophets from the Church but you'd need to compare them against experts in whatever field the prediction was being made in since the real world provides more information than they'd have in the coin flip case. The advantage of the coin flip is that there are fewer confounding variables more than anything else.