A few of my thoughts -
1. The IBM computer is not decentralized (in the coming years I believe the public will view decentralized systems with the same trust dynamic as they view the green SSL bar. Right not, it's not a big deal, but in the future, if a company isn't built on the blockchain, I believe they risk being left behind - perhaps in 5-7 years we will see this. I think sentiment is already showing signs of this starting to develop.)
2. The IBM quantum computer is a development in *hardware* and fundamental computer process.
3. Referring to point 2, I surmise that Golem (being software) can be adapted to quantum hardware in the future.
4. Golem, I think, will have great social interest in the future due to the fact that it will be one of the market providers in this new age of renting out computing power from regular desktops (which is rapidly being ushered in)
5. Like the article mentions, we're not quite there with quantum computers being competitive. Look at just that alone (time-frame) Golem has advantage.
My general intuition is that this poses little to no threat to Golem at this time.
1. The thing is not everything needs to be decentralized, something works better without decentralization. The decentralization hype will die down eventually.
2. If you can solve the computation with one hardware, why the need of using the whole world hashpower?
3. Only if consumers have needs to solve problem that needs multiple quantum computer processing power.
4. only if the price is competitive and indeed the speed is comparable to quantum computer.
5. GNT isn't fully functional as well, looks like this is a race of time.