Because it's inefficient. Google could buy a thousand biotech startups that are all doing innovative work trying to cure cancer and things like that, but that work has nothing to do with search engines or data centers, so it makes no sense to put it under the same roof. The same management staff isn't going to be effective in managing both types of business, all it's going to do is split their focus.
But give the money to the shareholders and the ones who want to can use it to invest in biotech, others can invest in electric cars or new types of power generation or energy storage or space exploration or whatever else.
Using the profits from search engines to fund cancer research and space exploration is great. Running those businesses inside the search engine company is pointless.
Yet it has a lot to do with Calico, another Alphabet company.
What if I want to invest in Calico but not Google, or vice versa? All it's doing is making the market less efficient.
Only they can't do it that way, because if you buy a share of Alphabet, basically the entire value is from Google and even if one of the little left-field bets triples in value it would have only marginal impact on the overall share price.
Meanwhile if you did want to do such a thing, the better solution would be to have a fund with the sole purpose of investing in high-risk biotech startups, and then invest in a slew of them to better diversify the risk. Then let retail investors buy shares in that entity, which is all biotech instead of being a tiny sliver of biotech but really mostly Google.
With all of the failed and failing products from Google acquisitions, is that really a great idea?
Firebase (acquired in 2014) seems to qualify as semi-recent given it's become a massive product line in a few years.
It adds value to their cloud, is responsive to actual customer desires, and generally seems like a bargain at the right stage (a Google-supported tool being more valuable than a third-party GCP tool, even with Google's questionable dedication to longevity).
Just don't make the IBM / MS mistake and think adding more PMs to previously successful company is a good idea.
There are an almost unlimited number of things to do with money. Even if you have lots of money, the number of good opportunities exceeds the amount of cash. It doesn't even have to be high-risk acquisitions - Just anything that will get a good return.
No mega company really has to have internal innovation. And I doubt the pretend internal startups ala skunkswork with an unlimited budget and extensive runways are a real replacement for innovation popping out of the marketplace among the bodies of a hundred other failed/acquihired startups.