I didn't say it was unimportant. I might say I wouldn't be concerned at looking at the individual monthly totals and trying to explain them.
In general, it's difficult to glean information, because any changes are only net. Without knowing either the number of approvals, or the number of additions within a period, I find it difficult to tease out truly useful information.
That's without even considering that there are cases that will never (theoretically) ever appear in the Inventory. i.e. the I-140 is approved after publication of the last Inventory and the I-485 is approved before the publication of the next Inventory.
My memory is a little hazy (since it was some months ago) and I am not going to repeat the exercise, but where there was sufficient information in the Trackitt case, I remember around a third of EB1C Trackitt approvals this year would never have appeared in the Inventory for that reason. I don't consider that figure very reliable, because only a subset of total cases had sufficient information to analyze. I did it more out of my own interest to just get an idea and I mention it only to illustrate the difficulties.
That's not to say someone else looking at the data might not have that aha moment - I've realized my shortcomings in trying to do so. And I have tried multiple times.
I would say the same about trying to gauge future EB3 demand from the PERM data. Another problem I have found intractable to date.
Enough wallowing in self pity. Sniff. :)
I think both those problems need new blood, with new enthusiasm to tackle them.