The point is, time stamps are unreliable because of the fact that "date accessed" means nothing and is often disabled (Vista, right?). Date Accessed could be pushed by virus scans, indexing, or any number of other things.
Well, it is just nice illustration of "unreliability by interpretation" :-)
They are useless if you do not use them properly.
There is no problem for Vista to turn Last Access time stamp on,
and there is no problem to create a script
conditionally using the timestamp if it is available.
I do not denying there can be improper timestamp pushing. It depends how SW is written.
But there are still many many files left untouched.
BTW, I recommend Avast AV, not modifying LastAccess while performing disk scan.
There is _no_ way to accurately determine what files are frequently accessed
based on time stamps. Since the other time stamps are totally useless (date created, date modified), they aren't used either.
LastAccess timestamp is aimed for negative, not positive selection.
There is easy way to determine files that definitely cannot be frequently accessed.
Sure, it will left some not frequently accessed files,
but not using the timestamp lefts ALL of them.
BTW, date accessed is not digital, but must be converted to digital by comparison,
because all decisions are digital by principle.
Note: I used your term digital, however I would prefer logical - boolean valueIf user cannot use info about file modification, he should not blame a timestamp.
Of course, it cannot be used to access frequency determination, it has other usage.
For example frequently modified files, in combination with creation date.
In opposite to LastAccess, it cannot not pushed.
One can use it for both positive and negative selection.
Even filecreation itself has its limited usage as an age of filesystem item ( not file content ).
Finally, if claiming timestamps say nothing, notice yourself that paths themselves does not say much.