Let us look at factors you have pointed out:
1. Social networks: Google has 3 positions in Alexa's Top 10 for the US market. Only Facebook and Twitter are the only two sites that are social networks on that list. Other than Twitter (protected accounts), Bing and Paypal, every other site on that list is crawl-able by Google. Even for Facebook, it is better to have public than private pages (6,330,000,000 pages in the Google index). *
It is very much possible for Facebook to be double the size that it is at and also for Google to grow at the same time. There are more people using more of the internet every year. They all need to connect with others and also find other information they are looking for. Growth for both Google and Facebook need not be mutually exclusive.
The most tangible growth-related metric of year-on-year growth for Google is still pretty impressive: http://investor.google.com/financial/tables.html Till a lot of the numbers start veering drastically off-course, I won't worry much about it.
2. Targeted searches: There will always be niche companies who do this over a million domains. I do not think Google would bother too much with it purely because of the clout they carry in the search market. If you have to get good non-organic traffic, you need to be indexable on Google these days. There is also the fact that Google would not want to dominate in everything related to search as it leaves them badly exposed to anti-trust issues.
3. Non-web searches: This is more of an input/interface matter. It has been possible to search through speech for a while on Android and it is something that is only bound to grow as speech recognition gets more and more better across all platforms. What matters is the quality of results and if competing platforms can provide better results from a non-Google platform, then Google would be in trouble. Parsing spoken text is different from delivering results for a query. So, Siri doing a good job of parsing what you are saying is different from Siri being able to serve results from a different search engine than Google.
If the majority of people who do speech-based searches get their results from Google, it will be something for Google to rejoice and not fear. In fact, I think Apple may have even done Google a favour with Siri. There is now going to be a race to be the best voice-based interface on non-iOS platforms as a result of Siri. There is also the increased awareness of this feature now due to Siri.
4. Information discovery: If I am not mistaken, Google already is the leader on this front. They layer a lot of additional information into search results (which, I am not a fan of) and there is nobody of a comparable size who is competing with them on it.
* Alexa is a flawed metric. I am using it as a comparative/indicative measure. Google's page count is broken; but they do have a LOT of pages on Facebook that is crawl-able.