I am not attempting to accuse you of any purposeful misrepresentation, but I do understand that empty blocks are not being counted in the chart buddy's renditions of how full are the blocks on an hourly basis.
Accordingly, leaving out blocks (whether a more accurate representation or not) could be the reason that blockchain.info is coming up with different numbers..
I invite you to run the numbers yourself. If you'll notice the times when the red line agrees with the solid green level, those are times when there have not been empty blocks and they do not affect it at all.
Let me turn it around a bit. One of the complaints the small blockers have with BIP101 is that despite there being a 75% of 1000 requirement for BIP101 to activate, random chance would mean that that could be achieved with much less than 75% of the hashrate. This is true. Likewise, even if on daily average, blocks appear to be much less than 100% full, random chance can (and has started to) mean that at random times, the blocks *are* full and users experience congestion. One full block is not really all that big a deal but we are starting to see, two, three and even four full blocks in a row. That's potentially a confidence destroying poor user experience. This is happening now.
Bear in mind, we *don't* have RBF, we *don't* have Lightning Network and we *don't* have segwit. What we do have is a fairly straight-line log plot that showed this was coming and warnings from several people, ranging from 2-5 years ago that this would have to be addressed.
I'm not getting into any arguments about small block or big block, and all I am saying is that I do not really understand the various chart buddy representations and clearly why they deviate from the depictions on blockchain.info. If the various representations of chartbuddy is not clear to people, then it becomes a lot more difficult to stay on the same page regarding what it is that we are arguing about (that is if we are arguing), exactly.