This is a good point. I think it's fair to say that the tiny subset of businesses VCs are interested in investing in are the ones where a well-timed influx of capital has a chance of allowing massive growth.
Any novel business that appears to have those characteristics is ripe for such a speculative investment, but most of the valuations are based on the assumption of massive growth.
As soon as the growth trajectory turns out to be less impressive, the valuation goes through the floor, and usually results in a loss for investors (even if the company can be sold).
But this is by design, since the investment was highly leveraged on the upside, it is also leveraged on the downside and the firm will often simply fail.
Doing a typical startup (less than two years from seed round to massive growth or failure) is a lot more like flipping a house than like starting a business with a proven business model.
This is why there is so much trendiness and so many me-too companies... They are like real estate developers who saw that McMansions sold for 2x investment a mile away and decide to rapidly build nearly identical ones hoping for the quick profit. The goal is entirely in the short term and the long term is irrelevant.
Google found a very powerful way to sell ads, and once it built the infrastructure to do so it was able to massively broaden its product offering and appear to be almost a modern conglomerate. But the fact remains that ad revenue is the vast majority of Google's income, and most of the rest of its businesses are (in comparison) failures that would never have continued to be funded by VC-style investors. Alphabet reflects the investors' realization of this and is a compromise resulting from Sergey and Larry having a lot more control than typical founders.
Github is interesting. I wish I could see its financials! I had hoped for a lot more in the product arena from Github, but who knows, maybe they still have most of the $100M.