That's only if you're trying to be competitive with Google et al as a business venture. If all you're going for is good enough, that's not indicative of your necessary costs. What's more, the FOSS movement has a long history of accepting these kinds of compromises.

I call this the "Retina watermark fallacy"—when you equate something not being the latest and greatest with being unacceptable. When Apple introduced the "Retina" Hi-DPI display for the iPhone 4, it was good. But what's more, it was supposed to show that everything else was junk. And yet, if you looked at Apple marketing materials from ~5 years prior, you could find breathless ad copy about their then-latest displays that were (necessarily) not Retina quality. That means either one of two things are true:

1. either Apple was selling unusable junk prior to the introduction of the Retina displays, and they managed to mistakenly convince themselves and everyone else that this stuff was acceptable when it actually wasn't, or

2. pre-Retina displays were good enough, and Hi-DPI displays are simply better

The truth is lies in 2.

So, the takeaway as I understood it from the original question would be whether or not the FOSS world could produce today a search engine on par with, let's say, 2002-era Google. (I remember 2002-era Google, and not only did it work, but it was good!)

You can't build 2002-era google today, it's 2017 and the web is a much more hostile place.

If there was any way the FOSS community could fund a reasonable search engine, I'd be happy to work at non-profit wages to make it happen. I don't see any way.