The flaw of morphometry is not in the technique but in its usage.
Used to classify bees that are bred using other selection metrics, e.g. docility, disease resistance, etc., then it's OK.
Used as a selection metric for breeding, it becomes useless as a classification tool. You end up with the wing pattern you select for... just in the same way that if you selected for proboscis length you'd end up with bees with short or long tongues, regardless of other characteristics.
IMHO it's been badly used for a few decades but hopefully people are waking up to this. It was an easy measurement, required little in the way of sophisticated equipment (so was readily accessible), and it appealed to the fiddling / data gathering mindset that afflicts some beekeepers - but there's a difference between data and information