Your strategy has the same problem - you have to predict how many last minute travellers there are too. If you just leave an unnecessary buffer then the strategy fails. The one of us who does a better job of putting exactly 100 bodies in 100 seats - none left over - is going to have a more profitable airline. That is somewhat core to the argument us pedant capitalists make.
It’s just the math works out so on the margin having even slightly better than a 1 in 4 chance of selling at 4x the prices is a net gain. Suppose their reserving 50 seats and have a 26% chance of selling all 50, only a 50% chance of selling 20 or more.
That means on 1/2 of flights their intentionally flying with 30 or more empty seats, but reserving even 49 seats rather than 50 loses them money.
How and or when exactly to you plan to sell those tickets?
Do it after the price hike and people will just wait.
Do it before and you need to see the future to know how many seats will eventually be sold otherwise you just traded a 400$ ticket for a 100$ ticket. It’s no more a viable option than suggesting it’s easy to win the lottery just pick the correct numbers.