Because nowadays any exponential growth is called a "bubble in forming". And because all stochastic exponential process eventually crash, or "underperform", it sounds really smart to "predict" a crash months or years in advance. Especially when using non-logarithmic axes, the predictions and the crash look really impressive.
You don't, however, often see these prophets of doom profiting from their predictions, because the exact timing of crashes isn't that easy at all.
Its actually generally a lot easier to profit buying a bubble than shorting it.There are a lot more points of increase than points of collapse.
Paulson made a fortune shorting the housing market,plenty of people predicted the collapse but Paulson picked the peak correctly.
The interesting thing is given his shocking trading performance since then he appears to have been lucky rather than smart.
Didier Sornette has some interesting theories(and dubious models) about bubbles,well worth a read.
You don't, however, often see these prophets of doom profiting from their predictions, because the exact timing of crashes isn't that easy at all.