The article suggest that it was neither. That rather it was an engineering decision: the product did not have staffing to be properly maintained, and rather than staff it up, it was easier to turn it down.
that's backwards - organisational/cultural/management decisions made it be under-staffed, which caused the code to rot, which led to the "oh, it's understaffed and the code rotted, let's just kill it".
I think there's very good reason to look past one blog post, people, come on.
If I am a higher-up executive in Google, I'm not going to much care what the numbers on RSS look like now, it takes less half a brain to understand that Google Reader has at least a CHANCE at chewing into the bottom line hard if it takes off -- even if those chances are slim. I'd kill it quickly as possible. If I can do so early, with little fanfare, even better.
They could be exploiting an undocumented/taboo feature of reality: simply telling people a good story can cause them to believe ("There is no reason at all to assume...") things that are not necessarily true.