My guess would be the more time you are on the item page would mean you are actually looking for it?
If I have, say, 5 possible infs in a batch I'm looking for all of them simultaneously.
I don't stand in reshop or dig through repacks and individually skip to each item as I search, especially if the items are in the same section (eg. If all 5 of my infs are in kitchen repacks). I skip to it only when I've found it or I have to double check my memory on the color/flavor/whatever.
So by limiting it to how long you're actively on the item page it can look like someone only looked for something for 2min when they were actively looking for 10-15+min while searching reshop, repacks, etc., for multiple items at a time.
Because, back to my example, if I find 4 of the 5 kitchen items in the repacks and have to inf the last it wouldn't count the full time I was looking in mpm. Because, most likely, I spent most of that 15min on the other items' pages. So all that time searching would read as 1-2min.
And the full time I was in the batch wouldn't show up so it looks like I never looked at all!
(Unless, of course, the item was wandable; in that case it would be more accurate since you actively have to be on the item page to wand for it)
But, of course, batch time is also an inaccurate metric. But at least it's inaccurate in a way that won't get people yelled at by management who don't understand why the time can appear so low.