I work with Flash and I don't think it's buggy or bloated - that's overblown. It's not perfect, but it's definitely much better than people paint it to be (specially people who don't work with it).
In the same vein, HTML5/canvas is unreliable both in terms of performance, adoption, and consistency among different environments. People like to think of "standards", but when it requires as much hacks as any other 'standard' work nowadays do, and when it's not even available in half the browsers out there, I wonder how can anyone think it's the golden goose people like to claim it is.
Depends if you're using Windows or not. On Mac and Linux, the performance and reliability are pretty bad (compared to other modern software, anyway).
We are looking at Flash solutions right now to solve that problem.
And all those people not using CSS layouts for non-pragmatic reasons actually held up progress because it stifled the market for tools (and skills and frameworks etc.) that would work in the new manner.
So good analogy, but I think it argues well for the opposite of what you intended.
But this will eventually boil down to a difference in opinion, and I agree with your point that it would be better for everyone if more people were quick to adopt technologies that are better in the long-run.
And vice versa. There's a nice table-based layout on the Google front page still!
Look here: http://www.google.com/advanced_search
Do you consider that mass of table usage to be for tabular data?
The debate right now is one side saying you should use "HTML5" because it's more open than Flash while Flash is buggy, bloated, and proprietary. The other side is saying "I need Flash to give my clients what they want." Just like people started using CSS over table-based layouts when it was pragmatic, people will stop using Flash when Flash stops being the best tool for the job. There's nothing wrong with that.