Much more has been written on the topic in the intervening years, including by Aaronson. I'd challenge cwillu's statement that "nothing" has changed, but I'm a D-Wave employee* so I don't tend to wade in. More specifically, I'm not terribly concerned with what "the public" thinks about our computers. Academic and industrial partners are much more reliable judges of quality.
* I do architecture, algorithms and circuit design; not marketing. That said, I'd encourage you to compare our marketing budget against those of IBM, Google, Microsoft, and the other big players in the field -- if the public is to be swayed, it'll follow the splashy marketing.
So long as they talk about the number of “qubits” their machines have without prominently clarifying that they're using a heterodox definition of “qubit” and “quantum computer”, I will continue to maintain that nothing has changed.
There is nothing heterodox about calling our qubits qubits, nor are our qubit counts deceptive. The limitation is on the algorithms that can be performed on our systems; something that we're extremely clear about in our publications. We describe our systems as adiabatic quantum computers, explicitly calling out the difference between our current products and gate-model quantum computers. And, despite not being explicitly designed for gate-model operation, we've recently observed Bell violations using novel control protocols. This wouldn't be possible without "orthodox" qubits. So whatever your beef is, the critique you've leveled at us above is inaccurate to the point of being misleading.
* I do architecture, algorithms and circuit design; not marketing. That said, I'd encourage you to compare our marketing budget against those of IBM, Google, Microsoft, and the other big players in the field -- if the public is to be swayed, it'll follow the splashy marketing.