A recent paper by Gusbin concludes that ΔI=3/2 terms in K→πη can substantially modify particle-mixing contributions to K+→π+π0. Using techniques which can successfully reproduce the experimental features of the K→ππ decays, we argue that this seems unlikely.