This isn't true at all. The effect of taxes compound just like anything else. E.g. if Warren Buffet had originally incorporated Berkshire as a Bermuda-based reinsurer, he'd be worth over $400 billion today. If he didn't structure it as an insurer at all he'd only be worth $20 billion.

Tax savings compound at whatever the long-term ROE of the corporate entity. Therefore tax arbitrage is most important for hyper growth companies.

Nobody is saying they don’t have any upside, but as always it’s the opportunity cost that kills you.

Point taken about diminishing marginal returns. But you can see that even taking that into account, the impact of lax tax planning for shareholders is humongous. (Plus many startups are in winner-take-all markets, and effectively have increasing returns to scale.)

The opportunity cost is real. Tax arbitrage will take some amount of bandwidth. But it often takes much less effort to raise cash by minimizing taxes than it does to raise additional equity funding. The former can mostly be done by the CFO in the background, whereas the latter requires the attention of the founders.

Further it assumes your capital constrained, which might apply to some fast growing business but if your the magical unicorn of a profitable fast growing growing company then capital is the easiest problem to solve. Worst case you might trade a little extra equity, but again that’s linear dilution often based on growth at the current rate.

PS: If your growing 10% per month it takes 6 years to go from 1 million per year in revenue to 1 billion per year in revenue. It’s very fast, but hardly instantaneous. Keeping that up for 12 years means 1 trillion in revenue per year which is extremely rare. Growth will slow eventually, and when it does there are a huge number of things to optimize.