Selling out and crashing servers
The modern day version of major product success has become system crashes & overloads. Mainstream media associates high-demand and success with website failures. Today's "sold out due to high demand" equivalent is an HTTP 500 error, no server response at all, or "system down due to high-demand" messages.
As this kind of messaging was abused in the physical/traditional consumer goods world, it can easily be abused online as well. Artificially choking supply by controlling output of a video game for example, can just as easily be done on the web.
There are fine lines between honest mistakes or miscalculations in how much production should occur, and consumer manipulation/deception, or technical shortcomings.
Today, the New York Times referenced MTV's servers crashing due to high demand of a particular ring tone. The average reader doesn't know whether that's because 100 people tried to download the ringtone, or because 1,000,000 attempts were made; but they assume it was a "large" number. I happen to know that MTV's network infrastructure is in the top tier for handling scale like this, so it certainly wasn't on the 100 end of the scale, but the point remains.
These kinds of capacity scenarios remind me of traditional power grids in which infrastructure is built out to handle peak loads, even though 364 days out of a year the grid runs at an average of 10% capacity (or something like that). Massive dollars are spent to handle that one day in the year when
everyone
has their air conditioner on.