On #1 - governance attacks could be a major concern if the circulating market cap of NDX is significantly lower than the TVL of the index pools.
An attacker would need to acquire at least 800k NDX to meet the quorum, assuming other holders were not active, or a larger amount to override the holders if they were active. Liquidity providers would have several days to withdraw due to the voting period and timelock, but not everyone actively monitors their funds, and assuming an NDX circulating mcap far below the tvl of the index tokens, it could potentially be profitable even if you could only attack the LPs who failed to notice the proposal.
On #2 - we do want more people to farm NDX to ensure there is sufficient liquidity for users who wish to use the index tokens as ETFs. I agree that providing a better basis for valuing NDX in the long-term (beyond speculation) could help with this. However, exit fees would not be sufficient to give NDX the kind of value it needs for #1. The primary value is, and almost certainly always will be, control of index deployment, category management and strategy approval.
Here are two other aspects of this to consider:
a) Typical index funds have no fees at all, and the biggest DeFi indices I know of (DPI & PieDAO DEFI pies) have streaming fees of 0.6-1%. Streaming fees are taken linearly from a pool’s market cap over the course of the year, though, rather than from token burns. This is not really an option for index pools at the moment, at least not without significant upgrades.
That being said, our index tokens are very different from index funds in the stock market, as they require some amount of meta-governance which NDX holders must pay gas for, and require regular upkeep calls which NDX holders will pay for in the form of NDX rewards to keepers. So it makes some sense to provide fees to the governance token holders.
b) It will discourage people from arbitraging the price of index tokens on other markets. The ability to do all-asset mints or burns with 0 fees allows people to arbitrage the price of an index token on an external market relative to the actual value of the underlying assets without massive price differences existing between the individual token pairs, as is required for swaps in the AMMs.
With 0 fees, this ensures that users can access liquidity at the correct price, without needing to mint/burn the index tokens, which is the expected behavior from an ETF. On the other hand, adding a small exit fee could help to protect LPs on other markets from some impermanent loss by raising the price differential necessary for profitable arbitrage.
I would normally suggest doing some analysis to determine the value this would provide, but considering that the current tokens are not being burned very much (since most people are just staking) there’s not really any data to base it on. I would suggest that if an exit fee is added, it be ~0.1-0.2%, as this is not so high that it will significantly discourage people from providing liquidity, but should be high enough to give a slight buffer to LPs on Uniswap and other markets, and generate some revenue to NDX governance. I do think we should be cautious though, as having zero fees is generally more attractive than even a small fee.