Your reasoning implies a Zeno paradox. If you half the level of the tax, then that increases the take, if you half it again this too increases the take, therefore the closer you get to to a zero rate of tax (by halving the rate an infinite number of times), the closer you approach an infinite level of taxes paid.
This is of course because reality is a lot more complicated than you are suggesting. I can think of two particular issues that matter for the purposes of this thread. The first is that other factors play a crucial role in the tax efficiencies and evasions you talk of. Primarily the discrepancy between capital gains tax and income tax. If a rich man gets paid in shares, he pays 18 to 28% cap gains on the profits. If he gets paid in cash he pays 45%, this provides a strong incentive to gain the system - one that only applies to the rich. But TPA are recommending a system of 30% and zero%, so the exact same incentive remains - again only the rich need apply. So it's lucky that the rich paying zero taxes will furnish us with infinite income, because many won't pay a red cent under that plan. Common sense says that ideally there shouldn't be a huge difference between cap gains and higher rate income taxes. The rich would pay more if income tax topped out at 40 and so did cap gains (subject to protection for pension funds of course) than they did with 50% income tax.
The other main issue is to do with right and wrong reasons for wanting to reduce a tax. A bad reason to reduce a tax is because people get away with not paying it - that only suggests a problem with the collections system. A good reason to reduce a tax might be that it is unfair - but that would inevitably be a subjective judgment. But this is a pragmatic business, so I will suggest that there is only one obviously good reason to reduce a tax, and that is if it has perverse effects - i.e. if a tax at a certain level can be shown to discourage people from working, investing, taking risks to increase capital etc....
But here's the thing. A 75% income tax for the high paid will discourage many but not all. A 65% tax will discourage significantly fewer, and so on. If an entrepreneur considers investing £100K to start a small business that might employ 3 people, he will do some maths first to decide if its worth the effort and risk. So he might lose his £100k entirely, or he might make it back with a million on top. If he is taxed at 75% he is wagering 100 against a return of 250, so he needs to be very confident of a maximum return. I'm sure a 30% tax would make him much more likely to take the investment plunge, but so would 35, or 40.
But what about those three guys he was going to hire? We know that a certain rate of taxes along with other calculations will decide whether their jobs exist. But they aren't wagering to win huge sums, they are getting paid 25K, and they have bills rather than risks against which to balance that. So their maths is completely different, the amount of tax that can be wrung out of them before it isn't worth turning up to work any more is completely unrelated to the entrepreneur's situation. So what is the objective value of billing at the same rate?
