A chamber-majority market and the individual seat markets that decide it are the same question at two levels of detail. Simulating the individual races — with a correlated national swing, because seats do not move independently — produces a majority probability that can be compared directly against the majority market's price. The two are traded by different crowds and drift apart regularly, especially when a handful of individual races reprice and the top-line market does not.
Markets on Polymarket where seat-count vs individual-race coherence is the natural play: