This narrative is indeed entirely speculative, fantastical and unsupported. It's also laced with factual and interpretational errors, one of which I wanted to highlight. This particular error demonstrates how ignorance and confirmation bias can combine to result in a stunningly incorrect conclusion.
The error I'm referring to concerns the supposed co-location of Knox and Guede at around 8.00-8.45pm. SA seems to base this entire theory on Guede's account of his whereabouts that evening and Knox's known cellphone location data. The trouble is, he is willfully misinterpreting Knox's cellphone location data to make his "argument". Here's why:
The Massei Report states that when Knox's phone received Lumumba's text message at 8.18pm, the phone was connected to the base station at Via dell’Aquila 5-Torre dell’Acquedotto sector 3. Now, let's put to one side for a moment the argument that this base station actually covered Sollecito's apartment*, and suppose that this base station does not provide the apartment with a signal. If that's the case, then it can be assumed that Knox was not in Sollecito's apartment at the time when Lumumba's text was sent to her phone.
But where might she have been? Well, we helpfully get a clue as to the coverage area of that particular base station in the Massei report, when it lists Knox's phone activity from the night before the murder (my bolding):
(Massei, p322, English trans)
This is interesting. It is basically saying that this particular base station provides coverage to "the heart of Perugia"'s historic centre", including "Via Ulisse Rocchi, Piazza Cavallotti, etc". Now, the historic centre of Perugia, and Piazza Cavallotti, are both a) within the old city walls, and b) pretty much due south from Sollecito's apartment. In fact, Piazza Cavallotti is quite some distance from Piazza Grimana, for example. I would post a Google Earth image to illustrate this, but the detail is hard to make out in smaller images posted here. But if you have Google Earth (or equivalent), you will easily be able to see the distances involved.
So if Massei is to be trusted on its information about the coverage zone of the base station which transmitted Lumumba's 8.18pm text message to Knox's phone, it means that Knox could actually have been situated anywhere within a large area inside the city walls. Yet SA apparently chooses to conclude that the base station data indicates that Knox must have been in the very near vicinity of Guede at this time. He places them both in the vicinity of Piazza Grimana:
(IIRC, SA actually ludicrously hyperbolises this in another post into something along the lines of: "It's proven that Knox and Guede were within feet of each other at this time"

)
But the cell tower data does not tell us anything of the sort: in fact, it actually suggests that Knox was much further into the centre of old Perugia at this time, rather than anywhere near to Piazza Grimana. I'm afraid this is nothing more than plain, old-fashioned confirmation bias and bad reasoning on behalf of the person who decided that the evidence showed that Knox and Guede were in very close proximity at 8.00-8.45pm that night.
* Even though some (including SA) say that Massei contains an error in linking this particular base station to Sollecito's apartment, the same "error" is actually repeated explicitly three times in the report (my bolding):
(Massei, p323, English trans)
Is it really a repeated error, or does the base station at Via dell’Aquila 5-Torre dell’Acquedotto sector 3 actually provide signal to Sollecito's apartment?