The way the authors have analyzed their data it tells us this is a promising drug. Had they done an intention to treat analysis very likely the opposite would've been the result.
The only thing that trial tells us is that if you do a bad trial you're as clueless as before.
I am not a doctor, just a physicist, so I know about experimental data, but of course not about this study in detail. But as several sources dealing with this study confirm, it doesn't really tell us anything. So really at best it can be seen as reason to conduct a proper study, but the data itself does not yield any conclusion.
There's a lot in medicine that goes into publication that has results not even borderline as meaningful as this. I would be careful to label this as "it doesn't really tell us anything". Especially since another study in China reached a similar conclusion.
a) The quality, or lack of quality, of other studies does not affect the quality of the discussed study.
b) The availability of another study with similar results, does not affect the quality of this study.
Erm, even poorly designed studies tell you if there's any efficacy to expect or not. And it leads to better studies afterwards.
Science is iterative.