If this were true, would we not see more truly hardware startups in existence? There are countless hardware products that have demand, yet go unfunded or are not attractive to VC's because of the large upfront costs or various market / IP concerns. Ever tried to develop and then manufacture an ASIC before?

Hardware is just that, "hard". Crowd funding, or crowd "pre-ordering" is just one approach to give these types of companies a fighting chance at marketing and developing a product. If it were not for the lack of project accountability and poor hardware track record, I think Kickstarter would be a very good solution towards some of these issues.

Unfortunately until this changes, I think backers might have a better chance at the casinos. At least you have fun losing your money there (usually)!

I think there are plenty of reasons one does not see more hardware startups, but I would argue that 'lack of funds' is more likely being used as an excuse to mask deeper problems.

One can expect bad things in any venture where the development costs exceed the expected return. If anything, this would be an argument for general purpose computing and not shouldering the cost of developing an ASIC.

If you would like to see a Kickstarter with better project accountability and a hardware focus I suggest looking at Christie Street: https://christiestreet.com