The reason is, sale cycles are so long and so different from anything else, that it doesn't fit with their view of the world. (And if you try to sell to governments without being aware of that, you are going to fail).
Short story, in-law works for major city parks department and as we walked through the park, she tells me that they have to buy certain trash cans that cost $10,000 each. These are not special high-tech trash cans, they are concrete ones that have stones in them. I know they are heavy duty but it would be a weekend project for some DIY'ers. Why are we letting firms get away with ripping us off (since we the taxpayers foot the bill)?
We're not a startup, but we built and maintain a web site which is paid for and used almost exclusively by local govt. it's been running for about 3 years now. Fortunately it's not part of our core business (we did it mostly as a favor to a friend).
In year 1 we quoted a pretty low number to build, and host the original site. Because of the price level, and because of some expertise we had, it didn't need to go to tender. [1]. We built it quickly, got it up just fine, everyone was happy (including the client) so we invoiced.
If memory serves it took about 12 to 18 months to actually get paid. (in 3 years we've never been paid by them in less than 12 months overdue). Our actual client works his socks off to push it through, but the basic process is excessively rule bound [2]. Since we're not morons the price in year 2 and 3 has risen dramatically to allow for the administrative hassle, and delay, in getting paid.
We've seen companies go to the wall because of this cash flow issue. However others I've polled get paid promptly, so I don't know the criteria that gets you onto the slow track or on the fast track.
So it's easy to look at a trash bin and ask how it can cost $10 0000, but you have to factor in the pre-sales and post-sales costs. These are substantial and could easily be 10 times the price of the product itself. Given the cost of pre-sales, and the likelyhood of not getting the job, many companies aren't interested, which in turn limits the level of supply.
You're right that the bins are of a higher standard to be more durable. And the only way to know they are more durable is to test. And testing costs money. They are also possibly designed to survive extreme events (like a bomb). When growing up in the 80's (in some countries prone to small terrorist devices) it was common to either have bomb-proof trash cans, or no trash cans at all. My point is that there's a specification for them that the manufacturer has to meet and both producing to that spec, and testing for it, costs money.
As the tax payer we are paying not just for trashcans but for the 100% transparency that goes with govt spending. No journalist ever wrote a story on how much money a govt employee saved by taking on more risk. When they do get creative, and it fails (as by definition at least some risky options will) then the media is quick to pounce. So don't be surprised when their primary goal becomes not-to-fail and not reducing-costs.
[1] [2] govts are exceptionaly risk adverse. Tenders are designed to reduce their risk to really low levels, and as risk has a value it drives up the eventual price. While you may spend $20 on a trash can, and just replace it if it breaks, they can't do that. Break enough trashcans and someone will run a story on how the public money is being wasted. This is not limited to govt though, when you are spending not-your-money your goal is not to save money, but to reduce blame coming back to you. Nobody ever got fired for buying ..... ?
Part of the risk aversion extends to the supply chain. You may be able to supply 10 cans of the approved standard this month, and the govt goes through all the hoops to make it happen. Then suddenly you "pivot" and they have to start it all over again. Mostly the humans involved in the decision making process aren't interested in extra work, so they attempt to reduce the risk by choosing a bigger, more guaranteed, channel of supply.
Thus disrupting govt spending, especially with a startup, would be exceptionally difficult because the appetite for risk on both sides are fundamentally opposite. And of course when you say "disrupt" you mean to "make cheaper" whereas when you are spending tax money, reducing costs is simply not a goal.
If you want to disrupt govt spending you need to first change media, and public perceptions of govt failures, and the cost of removing "all" risk.
Gov't spending / sales cycles are in many ways analogous to Enterprise spending / sales cycles.
Very similar risk / gain ratio for startups.
If you've passed the bar, it works very well against your future competitors. Much more so, compared to enterprise numbers.
There have been startups we didn't fund because they seemed like they'd be bad for the world.
Also, there's nothing wrong in pornography by itself, unless you happen to be an extreme religious person.
A bad business for the world would be something like Zynga.
Regulatory uncertainty can cause difficulties investing in certain sectors that might prove quite profitable.
Lastly, I would note that profitability as an isolated metric is not a perfect indicator of future success. I can find a link if people are interested, but in his latest book Robert Shiller discusses firms that created some special financial products in the late 90s. The products were effectively betting on black swan events happening at a certain time. They ensured a high annual IRR but then if the event didn't occur their investors were left holding the bag. What made the idea profitable for the firms was that LPs that invested in years 4-6 did so based on IRR, without examining things in detail and effectively overpaid for a piece.
What if my total revenue was $1mm, with a total market size representing about $10mm? Technically profitable, but a terrible investment for a VC.
Ultimately VCs want to see a clear path to $50-100mm, or they'll be reluctant to invest.
But, PG and YC seem to have already touched that...at least tangentially with Coinbase.