Yep, that chart is intentionally misleading. If you counted for Facebook / Instagram every individual viewer of each photo, I would imagine the scale dwarfs that of Snapchat.
As noted in the article linked: "...88% of snaps sent to one recipient."
If the KPCB chart were adjusted such that one send = one upload (certainly a better measure) I doubt the delta between Snapchat and Instagram changes much.
I disagree. The distribution is likely heavily weighted on the high end. If the last 5% is on average sent to 20-30 people, that would be 1-2x the amount of uploads as the 88% sent to a single recipient.
