Two identical snowflakes: statistical hokum?

Everyone’s heard the old assertion that no two snowflakes are alike. I’d like to weigh in with these observations:

1. What is a snowflake? Do you mean a fully-formed ice crystal made up predominantly of water molecules? What about those badly-formed crystals or those that aren’t fully developed? Aren’t they more likely to be identical?
2. All objects are made of atoms, electrons, quarks and so on. Thus, at an atomic level, snowflakes are all identical. Perhaps we’re just talking about the number of permutations of the atoms inside the snowflake… of which there are many.
3. There are something like 10²⁴ snowflakes that fall to earth each year: have you checked them all?

Thus it depends how closely you look to determine likeness. And at a quantum level, the act of looking too closely affects what you see anyway. So how can you be sure?

This is yet another one of those statistical anomalies where scientists have extrapolated results and given a likely outcome based on large sample sizes, even though nobody is quite sure what — scientifically — constitutes a snowflake.

Granted, the scientific rigour employed is better than the Man from Del Monte who — if the advert is to be believed — stands in a warehouse stacked full of oranges, plucks one from the pile, tastes it and indicates the whole batch is good enough for his product. Such shoddy statistical practice is why I don’t buy Del Monte orange juice :-)

A similar argument has been applied to DNA and genetic fingerprinting (snowflake theory, not orange juice theory). Apparently, the chance of me sharing my DNA with someone is on the order of one in a billion. Yeah, and there are nearly seven billion people on the planet so it’s likely I share my DNA with seven other people. Makes you feel all warm and fuzzy inside when the police batter your door in at 2am and claim you’ve murdered someone based on irrefutable DNA evidence doesn’t it?