This is "trickle down economics" you're repeating. That theory has failed to provide any empirical evidence.
1. Inflation is good, deflation is bad and out of order high inflation is bad. Unless you literally store cash, you're not taking money out of the system. It's either invested or lended. Your point #1 is completely false.
2. Rich people don't create the middle class, and create only a very small set of people wealthier... if any at all.
There are diminishing returns for society of having exceptionally rich people.
As an example: mathematically having a few people in a town earn more than 100x the rest will not boost the economy as much, as having a distributed rise in compensation.
In our current world what you pay for services and products is not at all dependent on your income. Which means that an average American software engineer earning 150k will still pay $1k for an iPhone, $200 for gardening services and so on. That is the same price anyone else is paying... This leads to diminishing wealth redistribution, as the denominator is how often a service or product is purchased.
Our household income is within the range of 500k, we consume the same products and services as our teacher neighbors at less than half that income.(they probably consume more, given that they have 3 kids)
If our household income goes to $1mil - we will not need to call the gardener more often, we will not eat more, we'll not need more haircuts, we will increase our expenses but not double them.
If our household income goes from $1mil to $10mil, there will be even lower increase in spending that would distribute the income...
Now if every house in our neighborhood got a $100k annual raise, the demand for services and products would go up 100 fold.
3. It's not a zero sum-game. It's a case of diminishing returns. It's interesting that you present a mass market product as an example, where wealth is produced by mass market, instead of an example of rich people making other people wealthy.