Of the two acquisitions where I had good visibility into the terms, in both cases the original Investors were cashed out at approximately 2x their total investment and the rest put into an earnout/retention package for the people that were coming on board. It was a haircut to be sure, but it wasn't actually a loss for anyone with preferred stock. I would have thought with smaller starting chunks that would be easier rather than harder than it was after the dot-com debacle. I certainly defer to your greater experience here as it is much more current than mine. I made the mistake of figuring those were more typical than I guess they actually are.
At face value, an investors buy shares at price X and the HR acquisition happens at value 2X. A naive conclusion is that (X - legal costs - time)= profits.
Do you mean that these acquisitions don't make enough to pay for the failed startups? That the cost associated with investing in that startup are more than the revenue? That something about typical deal terms makes (reported) valuations at buy and sell time meaningless?
What happens in practice is that the acquiring company effectively recaps the startup on the fly. This takes a bunch of different forms but a common method is that a big part of the purchase price takes the form of restricted stock grants or signing bonuses to the employees, as opposed to cash or stock that gets processed through the cap table.
From the acquiring company's standpoint, this is logical behavior because the acquirer wants the people to be well motivated to work hard at the big company, and doesn't care whether the investors get their money back or not.
But this has the effect of putting a startup's founders at cross incentives with their investors. It's very important for everyone to act like adults in that circumstance, which often but not always happens.
Of course acquirers can overdo this and burn their relationships with angels and VCs in the process.