There is no way to measure the value of protocols before Bitcoin
The point is not to compare the market cap of Bitcoin or Ethereum to the market cap of previous protocols at all. The point is precisely that the previous generation of protocols created immense value but most of it got extracted at the application layer while, as you said, there are many small but highly profitable Bitcoin businesses, but the bulk of the economic value is in Bitcoin itself, not in those.
Contrast that to the web where there is little economic value in the protocol and a lot of economic value in web applications.
100% on it also going the other way, as we've seen repeatedly whenever there is a hack at the application layer. The market cap of the protocol drops at a higher rate than the "actual loss", e.g. last week's Bitfinex hack of $70mm in BTC knocked 20% off the price of bitcoin.
Bitcoin's liquidity and network effects, in a nutshell. I don't think a competitor will be able to match Bitcoin on those fronts. And it's not just about the developer and technology community, it's about the media, governments, companies, investors, regulators, financial institutions and even regular people who are getting used to the name Bitcoin.
But, honestly, it doesn't matter. This stack is blockchain-independent. Whatever blockchain or currency ends up succeeding in the long run doesn't take away from the fundamentals of this stack: decentralization of data and processes, cost reduction, and consumer power.
I focus on the Bitcoin blockchain because that's what the market has chosen to build on top of because there are no better alternatives yet. We'll see what happens if and when they come (e.g. ethereum), but I worry that it'll be too late by then and the market won't take it.
I'll try to drill down to specifics in the series of posts that will follow this, but I wanted to set the general stage first.
To answer your question what's unsettling about this (and not in a bad way) that since so much of this stack is commoditized it becomes very hard to extract value where traditional software business do (20-50% cut rates, etc.) so it is much harder to be profitable, but at the same time it's much easier to bootstrap an application when you have access to an open pool of data.
I think it's an interesting dynamic that, if this turns out to be true, will really change the way software businesses work at the application layer. But I'm 2-3 posts away from doing a deep dive into that part of the stack – there's some things I'd like to go over first.
That's where overlay networks come in: a combination of a purpose-specific overlay network (maybe something like counterparty, or maybe something like Side Chains) to make up for Bitcoin's blockchain's deficiencies while taking advantage of its liquidity and network effects.
It's actually a Canadian startup with millions in funding, but this is not their product. It's the Kik messenger, which has a built-in browser. They provide this library to make making apps for their messenger easier.
That is actually /not/ very good for the economy at all. If you abruptly reduce spending to the absolute minimum, sure it sounds very idealistic, but you're grinding most economic activity to a halt, causing hundreds of millions of layoffs, and minimizing the job market.
I estimate that we'd have about 60-70% unemployment, and no jobs for those people to go to because, well, other people just aren't spending any money.
I used to have diplomatic immunity as well. Getting to ignore parking tickets and other fines, police officers treating you like royalty, no security checkpoints or lines at the airport (and a very comfy ambassador's lounge), complimentary upgrades to first class and not having to pay taxes (sales or otherwise) at the host country are the first that come to mind.
I was a little bummed to find out about Facebook's acquisition of Parse. Luckily, I came accross Deployd.com and this iOS matching iOS library: https://github.com/clooket/DeploydKit