On validation, keep the measurement condition consistent with the profile: if you built from M1, validate the printed target with M1. M2 may show lower ΔE2000 against CC24 because the reference data and UV handling are closer, not necessarily because the profile is better; it is a different measurement chain. For analytical validation, compare measured Lab with the profile's predicted Lab for the same RGB patches under the same condition, and use CC24 reference ΔE as a secondary check. Visual evaluation still matters. Don't choose M2 just to get a lower number.