There are a few issues with the Stats Insider ranking system, but they do take into account positional differences through a focus on impact as well as stats:
Stats Insider's AFL player ratings are derived through a meticulous process that takes into account various factors and statistical metrics. Similar to our AFL predictions for every game, these ratings are not subjective opinions but are based on objective data and analysis. Multiple criteria are considered to ensure a fair and accurate representation of a player's performance. By canvassing AFL media votes, coaches' votes, and match reports, our models have a much stronger idea of players who played important roles in each individual game. However, these aren't the only inputs; we also use a custom, fantasy-like scoring system to split tiebreakers and sort the players who have received a mention elsewhere to accurately rank every player in the AFL
Positional expectations could create a different issue. Stats Insider lists Dawson as a general defender. For a general defender he gets a heap of possessions, clearances and inside 50's, which probably helps him get to his ranking of 7th of 670 players. I wouldn't agree that he's the 7th best player in the league, but I do believe that he is in the top 55 of the c.550 players who played a game or more in 2023. He also made the All Australian team this year, so excluding him from the top 55 players (I assume based on just stats from your example) wouldn't make sense either (noting also that Stats Insider has 214 midfielders listed, 180 of which are in the top 550 players).
Similarly, Pedlar is listed as a midfielder, and it could be his expected possession count (and other mid-realted expected stats) that have him ranked way too low (in my opinion) at 391.
Stats Insider also uses data from prior seasons, which can understate improving players and overstate declining players. This also means that they list more players than the AFL use to create their rating system (which I've tried to resolve by adding the "Poor" rating for players outside of the top 550, a pretty blunt tool).
But all players are rated under the same basis and the anomalies would largely be expected to be evened out once they are generalised to apply to teams (not individual players) and using the AFL's Elite, Above Av, Av and Below Av ratings (rather than being more detailed like top 10 players, next 10 etc).
This is a trade-off to the issue that you noted in an earlier post, in that overlaying the AFL ratings over the Stats Insider rankings doesn't distinguish between someone who falls just inside Above Average (like Crouch did, by the way) who isn't significantly better than the player who fell just under the bar (like Sloane did). Yet, I would have given Crouch and Sloane those ratings myself (scraping into Above Average for Crouch, Average for Sloane). It also feels right if you look up the AFL ratings assigned to their key stats on our players page of the AFC website.
That's a lot of words to say the system is flawed, but so is every system we could use, but it's likely to be a good guide if used at a general enough level (eg, where each AFL team rates for elite, aboave average average etc).
If there's a better system, then I'm happy to use it (so long as it doesn't take a heap more time). I just haven't found one.
”
). 
