I agree that timing the market usually doesn't work. But it doesn't work in both ways. You are equally likely to miss or hit both good and bad days, with the same or similar impact on total return.
The author's colleague had a blog post on the topic, reasoning that missing either the best or worst was not a particularly interesting "what-if" since the best and worst days are both associated with the same bouts of higher volatility. He used the 200 day moving average to try and model the effect of 'missing' both.
It honestly didn't turn out so bad, but one takeaway is that backtesting (aside from not being a necessarily great projection) does not model personalities.
If one is equally likely to miss both good and bad days, but overall the market is up 10%+ in a year (repeated over decades), then money left in the market will double every 7 years or so (Rule of 72 [1]). In that way it is almost always preferable to invest and hold in index funds.
The S&P 500 has returned about 13.6% return per year from 2010-2020 [2], or doubling your money about every 5.5 years. If that rate continues or increases then of course it makes no sense to try and time the good or bad days.
Of course, some will argue that a pure index fund market won't be priced correctly as it is active/day traders who continually buy and sell to set a proper market rate. If everyone buys and holds like Bitcoin then there is no true reflection of the companies value in the stock. Those that would go under otherwise may be buoyed or bought out just for their stock value. There are many opinion articles out there on the topic, but so far it hasn't been observed at scale (I think).
The good and bad days are not equally likely to be equally good/bad, though. Long-term, the market tends to rise; the good days will outnumber the bad in either quantity or quality.
1. Market falls sharply.
2. General public panic and sell, while market timers double down.
3. Market falls further, market timers panic and sell.
4. Markets rebound sharply, with the above-mentioned people missing those good days.
The key assumption is that at least some really good days usually follow really bad days.