Are people seriously questioning how good HBM1 is purely on the basis that HBM2 is being worked on? They're working on HBM2 like they work on new architectures and die shrinks, technology marches on and standing around applauding yourself for your awesome tech is a recipe for future failure.
The fact HBM2 is in the pipeline has absolutely no bearing on how good HBM1 is, it's just the way of technology, no doubt they'll be planning HBM3 at this moment of time too.
I've seen no one question how good HBM1 is based on HBM2. Instead, using basic reading and comprehension skills, I can see that people are questioning the intial VRAM limitation of 4 GB HBM, and whether or not it is equatable to a larger amount of GDDR5 memory. Based on the fact that there are much larger amounts of HBM being produced in the near future, its not unreasonable to assume HBM doesn't mitigate GDDR5 VRAM capacity limitations.