I think his argument is that "paid" beats "free" because paid can support, whereas free is by definition relying on other revenue (advertising, other paid products) to support it.
When a paid product dies, it's (often) because the product wasn't doing well in the market. When a free product dies, it's because the company couldn't support it through their secondary revenue streams. The product may be the Best Thing EVARRR, but it's a casualty of failures elsewhere. The "bad" comment is likely because it's a lot easier to measure direct risks (i.e., risks associated with your product) than indirect ones (risks associated with your secondary business model, which likely isn't your area of expertise; if it was, you should be in business there).
My problem with the countless thousands of people who are taking this opportunity to gloat using that argument is that they basically are assuming their conclusion, since their argument goes something like this: "Paid services are less likely to shut down. Google Reader was not paid, therefore Reader was not less likely to shut down. Google Reader shut down, providing more evidence for this claim. We-told-you-so!"
But where does the original claim get its support? Are paid services less likely to shut down? We read on HN all the time about post-mortems of, yes, paid services. If we don't know that, then it's just a misleading post hoc argument (and in the counterfactual universe where Reader was a paid service, all the naysayers are now chorusing 'It's a Google service! You can't trust Google! We-told-you-so!').
Can you give some examples of recent paid services that were popular and were shut down? I think the reason people intuitively believe this is because a free service has to do two things to be successful: provide a good quality, useful service to people, and provide a way to make money tied to that user base. A paid service really just has to provide a good quality, useful service - since that is what the people are paying for.
> Can you give some examples of recent paid services that were popular and were shut down?
I find that hard, because I'm not a business person nor do I usually make the money-for-time tradeoff; hence I mostly only use free services and am mostly affected by free service shutdowns - in other words, selection bias makes my personal experiences unrepresentative. (Is this similar to other techies - like Hacker News users? I think it is.)
However, we can make the general observation that services only sometimes survive the business running them, and businesses shut down all the time. To give two personal examples of paid services I used that have stopped or will be stopping in the near future: Intrade has halted all trading during an investigation, and probably will wind up shutting down; Zeo Inc is shutting down as it looks for buyers of its assets, and while the web interface I've always used to export my data is not yet down, I don't give very high odds for it staying up for more than the next year.
In both cases, I paid good money to use them; it didn't save them.
When a paid product dies, it's (often) because the product wasn't doing well in the market. When a free product dies, it's because the company couldn't support it through their secondary revenue streams. The product may be the Best Thing EVARRR, but it's a casualty of failures elsewhere. The "bad" comment is likely because it's a lot easier to measure direct risks (i.e., risks associated with your product) than indirect ones (risks associated with your secondary business model, which likely isn't your area of expertise; if it was, you should be in business there).
Whether this applies to Google, I'm on the fence.