You realize simply repeating the same arguments doesn’t make them any stronger, right?
As to composite polls, what SeaShell21 is obviously referring to is a composite which leaves out your poll because, and I am speculating here, it is the outlier as no other poll ranks IMG as highly nor BG as lowly.
Earlier you were touting “quality over quantity’’ —-same is true with polling, hence excluding a clear outlier might actually make a composite more reliable rather than less—whether that outlier be your poll or, for example, cal-preps which you have argued has a bias toward Cali teams. Simply put, including bad data in a composite does not enhance reliability.