2.2 is definitely the most obvious indicator of this. The addition of three characters had major implications on the amount spritework that needed to be done, but it also had implications on the level design. However, 2.2 also placed a very strong emphasis on revamping levels and various assets to match a higher standard of quality, and so it could be argued that a lot of this workload would have been sustained regardless.
Quality control is again going to be a focus in 2.3, but whereas 2.2 can be considered to be "rennovating the house", 2.3 is more foundational in its focus, reviewing control schemes, physics, and abilities at player guidance matches the quality standards that our visuals strived for in the last patch. This means less in terms of graphical work and more in terms of coding, but the experimental nature of it means it's less certain what our total workload is going to be at the end of the day. When it's done(TM) is still as tried and true a benchmark as it's always been, but we're hopeful that the next major update won't take anywhere near as long as 2.2 did.