Something I try to continuously tell myself is that "It's still early days" because while it doesn't feel like it - I think it is.
Look at the massive shifts we've seen in social media use (Friendster > MySpace > Facebook + Twitter > Snap) - at near any point in their usage you could have reasonably said: "I don't see how anyone could overthrow one of these - they're too big"
There is so much future still to invent.
Just my observation and not being described accurately and clearly enough here. But I hope you get what I'm trying to say.
The difference with computing was on-going Moore's law (forever young), which now slows/ends. Worse, though they still shrink, the cheapest node remains at 28nm. Cheapness fuels revolution. Peak silicon has passed.
We could get a new technology (like fracking etc and peak oil), as Kurzweil suggests for the singularity.
I don't believe search is mature. We haven't seen a lot of innovations in web search for a while and what we know as web search is probably the best Google can do. It's probably not the best possible web search though. Just the best from Google.
The reason nobody is there to challenge Google is because noone has come up with a way to fund a web search engine other than through ads. Challenging Google probably also means challenging the whole ad industry. It's a rough climb.