Nokia makes the most technically accomplished phones in the world and it is not close. The current N-series has a five megapixel sensor with Zeiss optics, satellite positioning, wireless networking, and a processor capable of things nobody has written software for yet. Nothing from Motorola or Samsung or Sony Ericsson is in the same conversation on capability.
Nokia is also the largest handset manufacturer on earth by an enormous margin, and those two facts have almost nothing to do with each other.
The company sells something like a third of every phone bought anywhere. That share is not built on multimedia flagships. It is built on cheap, tough, monochrome-or-nearly handsets that cost twenty or thirty dollars, run for a week on a charge, survive being dropped repeatedly, and are the first telephone a great many people have ever owned. They ship in volumes that make the premium range a rounding error.
Nobody in the technology press writes about those, which is fair enough, and the consequence is that the coverage of Nokia is almost entirely about the part of Nokia that does not pay for the building.
What worries me is the part of the premium range that does not work, because the hardware is not the problem.
Take one of these phones and try to do something ordinary with it. Send a photograph you have just taken to somebody. Count the keypresses. Count the menus you pass through, and notice how many of them contain an item whose meaning you would have to already know. The capability is genuinely there. Getting at it requires a model of how the system is organised that most owners will never build, because building it is work and the phone is not a hobby.
So you get expensive devices where most of what was paid for goes unused. People carry a satellite receiver they have never switched on and a camera they only ever use in the default mode.
The reason this is dangerous rather than merely untidy is that hardware advantages are temporary and interface advantages are not. Everybody gets a five megapixel sensor eventually, because sensors are bought from a supplier. Nobody gets a coherent interface by purchasing one, because it has to be designed, and designing it means throwing away the accumulated structure that every existing user and every internal team is accustomed to.
Nokia is currently winning a race on the axis that is easiest to copy while carrying a weakness on the axis that is hardest to fix.
Being the largest manufacturer in the world makes that a comfortable position to sit in for a long time. It does not make it a safe one, and the cheap phones that fund everything will not protect the expensive ones if somebody arrives who has thought harder about the menus.