More masternodes means less reward per masternode as well as more total cost to run all the masternodes throughout the network.
That's only the reward "share" you're alluding to. The actual reward at an aggregate level is a constant and independent of the number of nodes. It's essentially the proportion of the chain mined at "zero difficulty". Thats the bottom line and you can slice it up any way you like within that overall constraint obviously. The actual number of nodes is not the problem, it's rather profitability dynamics between the two groups and their increasing asymmetry as price rises. It's simply an unstable model for all the reasons I outlined in previous posts.
Anyways, I guess one point I was trying to make was that if you were to reduce the percentage of rewards going to masternodes, you'd need to also reduce the collateral required.
I don't see how this follows. The collateral is arbitrary as it is. If running a masternode is more profitable than not running one then the network will get populated. As previously mentioned masternodes churn so lots of them have ALREADY been sold. It's in fact possible for the entire compliment of 5000 nodes to be sold and for the network to still be populated at a steady count of 5000 because they simply change hands at different times.
The problem is measuring returns in Dash. It looks profitable when in fact is isn't because the return needs to be measured in dollars taking into account the capital gain/loss of the entire holding. ROI in Dash would only be meaningful if we were a stablecoin.
Please answer this question that you continue to side step. Why, if someone is willing to spend so much money (and time) to mine, wouldn't they spend the same or even less money to just purchase the coin and sell it at a profit later?
Because mining is a business. It's income oriented whereas speculative trading is capital gain oriented. They compliment each other. Mining is a competitive thing so if nobody else is doing it there's enormous profit to be made for you. As more join in it reaches an equilibrium point of diminishing returns and that's your prevailing network hashrate level.
Your (and other's) claim is that by reducing the proportion of the supply that's mined we become more competitive and improve our store of value. I dispute this claim because I observe that the store of value is directly related to how much scarcity is baked into the blockchain. You can measure it because it's a simple cost calculation that's easily demonstrated using basic accounting. The price that the miner "pays" for the coin is different from the price that the masternode "pays" for their coin and therefore when the secondary sale occurs that price is baked into the accounting for that sale. If the miner sells below cost, they'll have discounted that price, part of which will still show up as a loss. The miner still paid the full price for the coin and it therefore represents the "opening price" for the block. On the other hand if a masternode sells, even at zero price they'll have broken even.
This is the asymmetry that's cancerous to marketcap growth IMO. You dismiss it by making recourse to anecdotal arguments such as "but masternodes don't sell". It doesn't matter. These anecdotal hypotheses are irrelevant to the valuation of the capital flows within the protocol because they're there and they don't change based on how they're denominated. Whether masternodes "sell" or not is irrelevant to the argument, what is relevant is how much of the blockchain new supply capital is wasted and how much is invested back into the chain.