I don't necessarily agree with the Google/Bing answer.
Search is a tool, not a use case. Expectations of tools and patterns are wideley different depending on the use case.
Google/Bing search is a use case where you have an open ended querry and the bold font communicate clearly what you could add. Because open ended querries often are exploratory, and you can search "anything", it is more important for Google to suggest your phrasing to be able to handle the input better. This is in fact a feature more important for them than you.
But! Searching for specific objects, it is not necessarily the suggestion that is most important (bolded and visually prominent). Over the years, designing both e-commerce search and desktop software tools, many users state that they want to know as soon as possible if what they search for exists and match. That is something very different from Google/Bing.
While Google/Bing is an open-ended querry, searching for specific objects could be quite the opposite.
There is no right or wrong here, I just want to advise against seeing all search interactions as the same use case and then compare it with Google. That is very unlikely to be applicable to scenarios in different types of experiences.
My source? Years of testing. Both solutions work, depending on what kind of experience you're designing for. The only certain answer I can give you is that search is more than Google. If search worked exactly as Google when I browse Taco bell restaurants on their website, that would probably be extremely confusing. :)
Are you helping the user phrase their search better to explore an extreme amount of possible roads that could lead to a relevant finding, or is the search deterministic and should let the user as soon as possible when they found what they were looking for?