A commonly quoted statistic, based on various usability studies, is that the majority of people tend to use only 5 percent of an office suite's capabilities. Vendors offering products that compete with Microsoft Office are especially quick to jump on this "common wisdom," with the reasoning being something like, "If we can do that 5 percent, then most users will be happy with our competing solution."
This is one of those ideas that sounds good in the abstract but falls apart in reality. The reason: Yes, there is a common 5 percent of capabilities that are regularly used, but, from there, most people use other features as well. And the ones I use are more than likely different than the ones you use. For example, I use the review and editing features in Microsoft Word, but seldom touch the advanced layout features. Someone else might do the exact opposite.
This means that offering 5 percent of Microsoft Office's features isn't enough to attract users. Prospective customers will want the features that they are accustomed to using--features that, in some cases, are critical to doing business. And, since these features are different depending on the individual user, a competitor will probably need to match well over 50 percent of the capabilities of an established office suite like Microsoft's to really play on the same ball field.
It is this problem that is currently facing the new Web 2.0-based alternatives, such as Google Apps Premier Edition. They all tend to do that core 5 percent of functionality just fine, but are way behind when it comes to meeting the specific productivity requirements that each user brings to the table.
This means that if Web 2.0 applications want to become more than convenient online emergency options and really challenge the established desktop products, they will need to become much fatter and feature-rich, possibly even moving toward offline capabilities that essentially transform the Web 2.0 applications into old-school desktop applications.
Which means that to defeat the traditional fat desktop applications, Web 2.0 products may need to become more like them.