Well, you have to keep in mind Opera is/was used in corporate settings in Europe as the default browser. Like how much of IE count was based on companies using base window images, which only had IE. That may explain the great dive in Opera usage before the blink version came out. Also every place will have a different percentage of opera users, but all in all the percents seem correct over the various sites.
The main problem I see with opera's usage is the version fragmentation. There usually was the current version and a few from the last version. Now it's Some from current blink version, some from old blink version, some from current presto, some from old presto. I have a feeling that fragmentation will keep getting worse since opera seems to have adopted google's agile SDLC process. Not sure if it's to sync with the development of blink, or just to hop on the fad bandwagon.
I don't think opera's main objective is browser penetration, or selling browsers to embedded device manufactures anymore. I have a feeling they're trying to switch to a service type monetization scheme. Where they have an online store where people get "opera max" or etc type services. (i have no idea what opera max or their other services do.)