It may not matter to you. It is a well written piece and relevant to HN. Some people enjoyed reading it and/or got something out of it so they upvoted it. This is how it works.
I would be a paying search engine subscriber, even at higher price points if the quality of the results was satisfactory and the following points were met:
(a) everything is over SSL,
(b) there are no advertisements,
(c) there is a strict pro-user privacy policy without keeping logs and user profiling.
I think the saying "If you are not paying for the product, you are the product" has a strong element of truth in it in most cases.
This has a strong element of "build it and they'll come."
It leaves out a large part of the problem of selling something. How will enough interested people discover the "simple web app"? If there are competitors (chances are there will be) and the app's page is in the 12th results page of Google and PPC keywords cost $5 per click, getting enough relevant traffic to ultimately reach 400 subscribers can be quite a lot of hard (and costly) work.
As an indication, if you look at point 5 in this survey http://webappsurvey2009.techcrunch.com/ you will see that ~50% of marketing sites of web apps got 0-1,000 visitors a month.
It can be done but don't just expect to "build it and they will come."
Also, a book that sums up good points about selling a web app (especially if you are a developer) is http://www.startupbook.net/