Every URL is kind of an API. Some returns HTML, some JSON.

If I write a custom browser that only loads images from web pages then I'm abusing the internet?

You're not abusing the web IMO, however, if you sell that browser there's a good case for contributory copyright infringement as you'd be commercial aiding others in making derivative works from the website owners content.

> however, if you sell that browser

That statement would be true for any webkit based paid browser.

"Every URL is kind of an API"

That is exactly what I meant, but better phrased - thanks.

Not an 'API' in its most commonly used definition - but a publicly accessible interface to his chord search algorithm nevertheless. Wouldn't you agree?

Yes, it’s called a “URL.”