I'm not an expert here at all and your point is well made. From what I've seen/heard the following apply:

1) It's quite clear to see from a flint flake if it's been made by being "hit" - so you can usually tell human worked flint flakes from "bits of rock"
2) The level of working/quality and the shape can give some indication of age.
3) The more bits of flint you have the more reliable the date is likely to be.

Not that different from pottery - it always amazes me when someone on TT gets given a small fragment of what looks like rock/pot/whatever and says "ah form 35, that'll 35-70 AD" :-) I've been waiting to see what the recent breakthroughs in scientific dating has done to prove/amend dates from this type of pottery dating but haven't seen anything yet.

Mac