There are about 1.1 million high school football players (not "millions" as the article's author says, but still above a million). To get 242 of them to take part in a survey is at the lower end of credibility as a representative sample: 242 of 1.1 million = 0.02% (that's one-fiftieth of one percent). That puts the risk of getting the wrong answers up around 8%. That's even before you start considering how balanced the sample is regionally, socially, and so on.
So you can look at that two ways. The more accurate way is, this chart has an 8% chance of being wrong. The other way, less accurate but still arguable, is that 8% of the information in this chart is wrong. Following that angle, out of every 12 teams listed, one is in the wrong place.
Think Bama is really stronger than this shows? You could be right. Think there's no way Nebraska comes out that well with modern high school players? Yep, you might be right there, too.
And that's even if the 242 who responded are a good mix regionally and in other social, economic, ethnic, etc. ways. If they're not, it could be much worse than just 8% off. For instance, if the respondents are all Catholics, because the writer advertised his survey in Catholic Times Digest, that would explain Notre Dame being at 15 instead of 30 or 40. If they're all from New York and Pennsylvania, that would explain the ACC and B10 being favored in the responses.
Most likely, there's a regional bias built into the survey. That's hard to get right. But we'll never know, because the author doesn't report on the demographics of his sample.
Bottom line: the survey is interesting in the way a Reader's Digest article is interesting: wouldn't bank on it being right, but good fluff.