Well, here's an analogy, In semicondcutor chip design field, a design has made and prototype, the first silicon come back and tested, it works. But later when try for masss production the yield is only 40%.
The designer engineer complain it must be the process issue, therefore the process engineer should fix it. Then process engineer counter back hey your design works but the margins are too low therefore yield is low.
A design works doesn't mean anything, the most important aspect is how much margin you have above the spec. Because variance of manufacturing process can throw a weak design, with little margin off easily. A design with good margin will buffer against those manufacturing variance. hence good yield.
majority of time, if the yield is low, the design engineer has to modify the design to introduce more redundancy or margin for the chip. That usually solves the probelm.
So, that bring back the questions, how much margin does the WS10A design has? All indications seem like it's weak design with little margin.
Like the chip design case, I think this WS10A has to be modified.
Another option is wait for WS-15 which has designed with higher spec, so the WS-15 design would easily meet WS-10A critieria, therefore use the overkill of the superior design of WS-15 to generate good yield. (the WS-15 design should give very good margin when stacking against the WS-10A spec)
