I'm curious about the tail emission and transaction fees after major coin issuance. What is the minimum value that 1 XMR must sustain as an incentive for miners to continue maintaining network security?
This is kind of a subjective question. Currently the block reward is ~ 10.4 XMR per block or 5.2 XMR per minute. When the tail emission kicks in, the block reward will be 0.6 XMR per block (assuming 2 minute blocks) or 0.3 XMR per minute. If we further assume miners are currently mining at an equilibrium then the price should be approximately 17 times as high to sustain the current hashrate. However, hashrate usually lags price and thus it could be that miners are still mining at a significant profit instead of an equilibrium. In addition, it could be that miners are further optimized in the future, which results in an increase in hashrate for the same amount of cost to a miner.
Thanks to the tail emission, in order for miners to continue to be incentivized, Monero must merely have a price. It honestly doesn't matter what that price is.
The security of the network scales mostly linearly with the price of Monero, but so does the need for security. If Monero has a market cap of $1 thousand, there really isn't much need for network security, but the 432 new coins every day will still have a tiny net worth, and thus someone will be incentivized to mine it, and maintain the tiny network. If it has a market cap of $1 million, there would be 1000x the incentive, and 1000x the need for security. $1 billion --> 1000x the incentive of a $1 million cap, and have 1000x the need for security. And so on.
Thanks to proof of work and a tail emission, as long as there is any incentive, the network security will match that incentive. If Monero is used and can be traded, there is incentive equal to the potential network security.
Since it also depends on the transaction fees being paid - and hence how much the network is being used, and prices of electricity gpucs/cpus asics at the time. I'm a little tempted to say this is an unanswerable question since it is a function of many difficult to forecast variables