At every moment of disruption in technology people saying that the new technology doesn't replace the incumbent. By definition disruptive technologies are less functional and inadequate "replacements".
First, folks tend to talk about all the things that the new technology can't do that the old one does do. In the Steve Jobs interview at All Things D referenced in the comments, he goes on to talk about how software needs to get written--"it is just software" he says. In the near term history we have seen this same dynamic in the advent of the GUI relative to CUI or in the way browser/HTML subsumed the GUI client-server apps. People are writing more code all the time that is "mobile only" even if some of it reinvents or reimagines the desktop/laptop world. I was struck by Adobe's recent developer conference where they showed many mobile apps. As an always aspiring photog we can see how the field is transitioning.
Second, people tend to underestimate the way that new tools, as ineffective as they are, drive changes in the very definition of work. Said another way, people forget that tools can also define the work and jobs people have. It isn't like work was always "mail around a 10MB presentation before the meeting". In fact a long time ago meeting agendas were typed out in courier by a typist -- that job was defined by the Selectric. The tools that created presentations, attachments, and follow up email defined a style of working. While we're reading all this, the exponential rise of mobile is changing what it means to work--to go to a meeting, to collaborate, to decide, to create, etc.
What is so fascinating about this transition is that we might be seeing a divide where creators of tools will use different tools, at least for some time, than the masses that use tools. Let's not project the needs of developers on to the whole space. We might reach a point where different tools are needed. Two years ago I might have said this applies to a lot of fields, but the rapid rise of mobile and tablet based software for many things is making that argument weak. Cash registers, MRI machines, video annotation, and more are all scenarios I have seen recently where one might have said "needs a real OS" or "this need sa full PC". As with the the idea of underestimating software, our own desire to find an anchor pushes us to view things through a lens where our own work doesn't change.
All of this is happening. In parts of the world they are skipping over PCs (Africa and China). Everyone is seeing their time in front of a screen go up enormous amounts and most of that is additive, but for many there is a substitute effect. This doesn't happen overnight or for everyone. TO deny it though is to deny the very changes that led to supporting the idea that the mouse, overlapping windows, and color once displaced other technologies where people said those were not substitutes for the speed, efficiency, or capabilities of what was in use.