Indeed... deluxes are getting smaller and lighter to combat inflation, so comparing toys from the same size class from several years apart will be flawed. Comparing among one series or year is more scientific because it eliminates more variables that aren't the focus of the study. If you then find one significantly larger or smaller in THAT sample, it is indeed an anomaly and warrants further discussion or investigation to find out why.

(I would normally focus on weights when comparing toys or looking at size-class shrinkage, as the altmode and kibble can make something deceptively bigger or smaller.)