Actually, this is misleading.
Having all your teams ranked inside your conference means you move up 1.5 spots NUMEROUS times...so while there isn't data to show that huge poll jumps happen...there is data that they didn't look at to show how often ranked teams are played.
See, the SEC can play 4 or 5 ranked opponents in their standard conference games...teams like Nebraska may only play 2 if they're lucky. 1.5 spots x 4 or 5 is a pretty good thing to have the option of and conferences outside of the SEC don't have that option thanks to preseason rankings etc. for the SEC.
The real thing we should be paying attention to isn't the number of spots teams jump...but rather...the number of games versus ranked opponents that give the opportunity for jumping. If the SEC switched to a schedule like that of the other Power 5 conferences to match them, the SEC 'bias' would probably go away. In other words, the system is broken and showing stats like those in this article definitely DO NOT prove or disprove anything.