In this case, if you limit the acope to those receiving document requests after June 13, and you apply a linear regression against the dates, and you add a best fit line, you can see what the average time from document request to landing is currently 3 months.
Now, I agree that the data may be inconsistent, but if you accept a range, 1-4 months, pretty much all the cases are landed.
CIC's call centre reps are not always correct. They don't want to do anything if you are within the posted time (which is wrong also, from CIC's own open statistics). But returning to the same dataset I mentioned above, you can see AOR to landing (no SA, no AIP, no DM) as of late. And it is possible for them to fo this, given their SIR program.
You don't have to accept this at all. But I developed civil engineering applications that, if you have ever used a road or flown in a plane, you've inadvertently benefitted from. That applications is used worldwide and by the World Bank.
Just like the mess of data in the spreadsheet, applications I used to build would take data points collected by road personnel, verify that the data is in line with previously accepted runs, and would perform evalution on that data for quality and roughness. That was finding order in 18million data points, not a few hundred.
I provide that so you may understand where my opinon comes from.