Namada isn’t limited to using NAM to pay for transaction gas fees–we can use governance proposals to enable our validators to accept arbitrary tokens to pay the gas fees. Personally, I think we should do this, ideally for as many tokens as possible.
Primary reason for accepting non-native fee tokens
It makes the user experience much better. For example, if I deposit OSMO with Namada, I can immediately use Namada without having to acquire NAM (which can frustrate the user experience). I think we need to make Namada easy and desirable to use for the immediate future, and we cannot afford to impose frictions.
Primary reason against accepting non-native fee tokens
The only reason I can see against this is that forcing the user to pay with NAM requires them to acquire NAM, increasing the demand for NAM.
Counterpoint
I imagine that people will ask: then why hold NAM? For staking. NAM stakers will capture the transaction fees and will vote to decide the future of Namada.
My counterpoint is that until Namada block space is saturated (ie. there’s competition and thus a market for processing Namada transactions), NAM fee value will be insignificant and thus won’t drive demand for NAM in a meaningful way.
Future considerations
Fees have historically been a means of preventing the chain from being spammed with useless transactions that prevent actual use. If we think fees will be a meaningful way to capture value needed to sustain the network, there are things we can do to make NAM the preferred fee token, like making it 10x cheaper to use NAM. If NAM demand is important, then perhaps there could be a way to automatically convert non-native tokens to NAM within the system, so that NAM is still the unit of account.