Well, if your attendance is consistently above the line then no, you wouldn't decrease your prices to the optimal level.
But the line is simply based on the average of all the sampled games where teams had that indicated ticket price. In that sense it's a pretty conservative estimate of optimal ticket pricing. Obviously the attendance is dependent on everything in the fan survey, and also whether you just promoted/demoted, so there will be cases when you should choose a price point higher or lower than the "optimal" cost.
I found it very useful for setting prices as I promoted into a new league this season. In this case, I looked at the higher end of the actual attendance figures, and set my prices accordingly. Also, in the past I've used it when I know will upgrade my arena in a time frame sooner than the next ticket adjustment price. In that case, I've set my prices so that I will be near sell out levels once the upgrade is completed.
As for the r2, that is the correlation coefficient for the fit of the data to the equation at the left of the (r2). As for 1.500, it's probably not meant as anything important.