A couple of points I can reply to :
1) From the climate scenarios we develop in CRU there is no evidence that there will be any increase in tropical storms. In the area with the best data - the tropical Atlantic, there has been a reduction in both the numbers and the severity of Atlantic Hurricanes over the last 50 years. There has been a lot of US work on this subject. Although only applying to the US area, the work shows that damage ( when normalized to a common point in $'s) and lives lost have both reduced. Claims are much higher because of greater insured areas and the much greater population living in affected areas (particularly in Florida).
2) A recent paper in Climatic Change by S. Ungar, 1999 called 'Is strange weather in the air ? A study of US National Network News coverage of extreme weather events' shows that there hasn't been an increase in extreme events reporting (global areas) since the 1960s.
Most lay people beleive there has been an increase because the pictures make news stories. In the past there were reports but no pictures. The media also always like an explaination for an extreme, so the greenhouse effect or ENSO often gets the blame. There have, however, been few studies which have attempted to look at extreme events on a continental scale to se whether they have been increasing or decreasing in frequency.