It would still be great for a very early prototype or proof of concept, but the choice of parts leads to some limitations as a prototyping platform because it would be hard to graduate from it. Most people would not be able to use a similar part in a “production” device. If you have enough physical space you could just use a Raspberry Pi, but then you would need to be confident that you could always buy it in sufficient quantities.
One great thing about Arduino is that even though it’s really easy to prototype electronics, the AVR microcontroller on the board is a “real”—not just hobbyist—part used in millions of products, it’s always easy to buy in any quantity from one to tens of thousands, and you’re never stuck in the prototyping platform. (Sourcing hundreds of thousands or millions of something is probably always hard.) I’ve created a half-dozen devices that were prototyped on an Arduino and then moved to a custom PCB, and it was always an easy and pleasant experience. If you stick to 16 MHz you don’t even need to change any settings, let alone code.
Obviously an Arduino would be underpowered for the kinds of things you would use a Raspberry Pi for, but for prototyping one might consider more expensive alternatives such as Chumby Hacker Board or BeagleBoard.