He just arbitrarily adjusts the model based on what he thinks he did wrong, so again, there's no baseline and it's no different than if I created my own model. It's not based on some established precedent for winning, but only on recent results, it's always based on what he thinks it should be.
I don't think he's been more right than traditional NHL analysts at all, and the fact he was too low on the Islanders twice in a row (when they did well) and then was too high on them last season (when they did poorly) should tell you everything you need to know about his "models." He doesn't have faith in the model, he's adjusting it to fit what he thinks will happen.
They both would've gotten $5M more on the open market, yes. Bailey actually took a discount at the time he signed that deal. Palmieri hit the open market and was brought back. If you're basing your "overpaid" claim on a single season, I guess, but it's a really poor way of viewing things.
They lack top end talent, which is what I said earlier. Having more middle 6 guys than the opposition doesn't suddenly raise them all to McDavid levels. The COVID bout really did a number on them as well, it guaranteed half a dozen losses while the rest of the NHL didn't have that. Late in the season the team was actually ahead of Washington in regulation wins, but were double digit points behind Washington in the standings. That's because they can't win in overtime when other teams put their top tier guys against the Islanders' overabundance of middle six guys and no game breaking talent.