Well, first, guess let's assume that's after-tax money. If that's before-tax, you end up losing money. Secondly, let's be honest. That's not really a $51M base salary. That's a $1M base salary with a $50M signing bonus that you can't spend. Because if it goes south and you need to pay the $50M back, you better not have spent it buying some nice house and other stuff. What sane person would risk the chance of going bankrupt to pay back a salary after they received the money? The smart decision is to just not touch the money until the clause is up. So put into perspective, it's not a $51M base salary at all.
Besides that, the issue is that these decisions aren't made in a vacuum, and the rewards are too short-term to make it really viable. The CEO will consider the offer and compare it to other opportunities out there. Free agents always have options if they're that good. It happens in sports, and it happens in business. So you have candidates comparing options, and those options drive up the price. If the price is driven up because options exist, any sane person would be a fool to not take the best price if the opportunity also fits right. "Sorry, your offer is the best offer, but I'm going to turn it down because this other company wants me to give back 98% of my salary if I fail. That's more my cup of tea." I don't deny there can be non-monetary factors that may be more important than money in these decisions. But I'm talking about an all-else-equal decision. Especially when it's a turnaround situation. Nobody wants a turnaround job unless they enjoy pain. The money is to compensate for that pain. Nortel had a really tough time finding its last couple of CEOs. In Canada, it was known as the CEO job that nobody wanted. How do you attract someone for that kind of job unless you guarantee them nice money?
Finally, the market does reward for the short-term. When IBM made the big switch to go all-in into services with the purchase of PwC Consulting, mass global hiring for labour-intensive work, and so on, they opened up a huge profit tap. Palmisano was hugely rewarded for growing this decade-long windfall. He parachuted into retirement with a very nice package. Never mind that the strategy worked for only a decade and created a big mess of financial misallocation that his successor would need to clean up. Should he have been able to predict that IT was changing and that this huge IT services workforce that he had hired would turn into a huge albatross that Rometty would need to slay in order to ensure IBM remains a relevant company? Maybe. Is it fair that Rometty is left dealing with the albatross while Palmisano is happily and richly retired? Yeah, Rometty was part of executing that services strategy as a senior executive before she became CEO (in fact, she was apparently the one responsible for integrating the PwC acquisition), but the question stands. Under the proposed scheme, why is Palmisano rewarded and Rometty not? Rometty is dealing with the fallout of a changing industry overshadowing what made perfect sense to Palmisano at the time. They were hit by technological trends that they didn't predict well.
Compensation is a complex subject, especially at the executive level.