There is a handful of nice reading comments to this article. Sure it's not easy to predict what the future of Linux will be in the desktop, with Google.
IMO what is granted for sure is Linux wont never disappear, regarding the consistent stack of market share for so many years.
Also IMO the fluctuation in the Linux user base will not depend on what else can be done from inside to increase it's adoption, because today Linux already offer almost everything a PC user may need.
I don't really see that Adobe could be the magic key. Does someone help me to estimate how many new users they could bring aboard?
Office is a different problem. Both M$ and Linux average Office users perhaps only use a small part of the total potential that both products offer. This mean that the problem is not related to any quality differences among both. It's related to their compatibility. Nothing can be done here as it depends on external regulatory mandatory standards that may well never happen to be seen.
All together, the unique solution is to change the global user preferences regarding the utilization of a few productivity tools . This is marketing and cost a fortune. The Linux ecosystem hasn't this potential. Even if it had, it's a task that takes years and has a risk to fail.
Google (or any other top tech entity) is capable of that. But Google is a commercial company that exists to make profits. They have to carefully evaluate how much revenue they generate for each $ they spend. Does someone guess what the bottom line will be for them? Well.. so far I'm a happy Linux user. No matter what the answer is, it will not affect me too much. :)