Have people ever wonder why Google decides to include Flash? (Aside from trying to one-up Apple).
I think there are real technical reasons why Google ships Flash along with Chrome. It recognizes the inherent slowness of HTML ratification and Flash is a logical choice to complement the shortcomings of HTML.
The strength of HTML is that it's an open format. HTML mark ups facilitate semantic meanings to its content which allows easier parsing for search engines (and ultimately benefiting the users). Most of the the web is heavy on information (news, blogs, review sites, weather, etc), and this is where HTML excels.
However, HTML will always the lowest common denominator technology, and will always fall under the mercy of the many different implementations (browsers). Web developers will always have to spend significant amount of time to cross-test major browsers and write code with graceful degradation in mind. This is very cumbersome if you are trying to push boundaries.
The soon-to-be ubiquitous HTML 5 specs introduces many welcomed features, but it won't overcome the inherent nature of HTML - that is, the actual age of an HTML spec is in span of half-a-decade and more, where as, plugins are revised in terms of years.
This is why there are plugin specs within the HTML specs. The plugins are in essence, incubators of ideas. They are the innovators and most importantly, they are there to address the short comings of the current HTML specs.
We all know the short coming of Flash (security and performance problems on second-tier platforms), but let us at least recognize some of the positives:
It had the first successful implementation of vector drawings for web - now the open implementation (SVG) is finally taking off with Microsoft on-borad.
It single handedly brought ubiquitous video to the web, where all other giants (Microsoft, Apple, Real Media) had failed. Video tag is finally coming to HTML5, which will finally have most of the basic features (There is no support for multi-cast, streaming and etc).
Flash's text-layout framework brings desktop publishing text support to the web. With support flowing between different text flows, vertical text and etc.
Many of the current Flash functions will be replaced by HTML5 (videos, simple charts, audio, simple slideshows). But Flash is already heading towards directions that are not explored. For example, Flash's performance is acceptable enough that there are usable 3D engines, something that VRML failed miserably in the 90s. Browsers' canvas performance is simply not there yet. WebGL is also years and years away. Browsers are now looking to hardware accelerations for rendering, where Flash has basic implementations for ages already.
The truth is that since version 9, Flash is gearing towards rich internet applications where semantic meanings are not that important (the application is not the content, it aggregates content). Flash provides fast start up time, write-once-run-anywhere, good-enough-performance, and maturing IDE platforms (open source such Flex based on Eclipse).
While JavaScript engines are gaining speed, ActionScript 3 is still especially competitive - it's hard to beat pre-compiled code. The biggest strength comes from that developers usually only need to test against one version Flash. JS+HTML+CSS+Different Browsers is especially cumbersome.
Many people falsely believe that Flash is completely proprietary. The swf spec is open to all, there are many open source tools on compiling both AS2 and AS3 swfs. The Flex framework is open, and runs on Eclipse. There is a open source Flash runtime called Gnash (poor performance and AS2 only though). You don't need to Flash Pro to make swf files, though it is the best for making animations.
It isn't open as in W3C open, where members can veto features and etc, and the documentations may not be the best, but it certainly is far from what many critics calls as proprietary.
The challenges for Adobe on Flash is mostly the same as always, how it can continue to innovate, but this time - it needs to address the problem of developing on limited mobile platforms (perhaps performance profiler for mobile devices?) while improving security and performance on second tier platforms (Mac OS X and Linux).
I believe Google doesn't think that HTML5 will replace Flash, it's thinking about how HTML5 will be bolstered by Flash. The future is HTML5, but there will be always plugins (such as Flash) pushing the boundaries.
Apple has also made Flash work better by isolating plugins so they can not crash the browser...
