Yeah, in the last two days there is increasing evidence that there's something else going on in at least some patients beyond "pneumonia resulting in ARDS"
That tweet is interesting. Some folks are starting the theorize that many of the pneumonia (especially, bilateral) diagnoses based on CXR might be mistaken.
The most interesting hypothesis I have encountered (way too little evidence yet to do anything but study it) is that the virus is directly interfering with haemoglobin saturation, and that the abnormal CXR is actually looking at tissue damage caused by haemoglobin degradation somewhat akin to that found in Malaria patients.
That hypothesis intends to explain the apparent efficacy of chloroquine, magnified when taken by patients early in the disease, by hypothesizing a similar mechanism of action in COVID as in Malaria -- not by being antiviral, but by inhibiting heme polymerase.
The apparent antiviral properties of the treatment are then merely keeping the patient healthy and lung-damage-free while the immune system does its thing.
If that were true, then treating COVID patients for pneumonia could be highly counterproductive.
Of course, even if the hypothesis is correct, there might be some COVID patients who do develop a bacterial secondary pneumonia -- that is an uncommon-but-not-rare complication of all respiratory illnesses. That might also explain why Azithromycin seems to make Chloroquine more effective -- each drug helping a different subset of patients, rather than actually as a combined effect.
As I say, though, there is nowhere near enough evidence yet for this hypothesis -- but it is pretty interesting. It is probably worth a study or two to evaluate.
It is a neat hypothesis in that it takes a wide variety of disparate and otherwise apparently contradictory datapoints from clinical experience and weaves them together into a coherent narrative.
Of course, that it makes a good narrative is in no way evidence for it's truth, no matter how much we humans love good narratives.