You pluralized the thread title with the word "analysts" when the articles you linked clearly indicate this is the spreadsheet modeling of one single analyst.
What is also clear in this one analyst's modeling is that she assumes desktop sales will continue to trend down, in other words she has to make an assumption that desktops will be even worse in Q4 as they are now so as to make a large artificially projected revenue shortfall to which Intel must then make up by selling even more non-desktop processors.
What we have here is a classic example of
confirmation bias. You clearly want Intel to do badly, and any tidbit of information you can glean from the internet to support this desire is picked up by your eyes and immediately taken as "the truth!".
What do you have here? A single analyst who has already gone out on a limb to rate Intel as "underperforming" in the past, and who is now trying to justify that rating by concocting hypothetical revenue shortfall models to save face.
You can find yourself a single analyst to support just about anything. I wouldn't be surprised if you could find at least one analyst out there who thinks the moon is made of cheese in their professional opinion.
...
You want an intelligent assessment of Intel's revenue model, not a one-off solitary calculation done by an analyst who at best can be painted as "confused".