In its current state gas is rather irrelevant (as in nearly free, which logically makes it highly competitive) on the Fusion blockchain. A majority of txs, are either regular txs or swaps/functions of AnySwap. Neither of which currently costs much gas (often only 1 gwei =0.000000001 FSN), very much in contrast to the same actions on Ethereum, which are quite costly.
But for the future of time value , it's actually important for the gas price to become higher (or at least the total amount of gas used per block. Ideally that should be able to measure up to 2.5 FSN [the current block reward], for a predictable time value of FSN to exist).
By predictability of time value I mean the ability to determine how much interest FSN can yield over the course of a a month, a year or two years. This is important, because the easier this is to determine or at least guess, the easier it is to trade with TL assets spanning more vast periods of time, which in turn is important in making TL txs popular (which we will see is a likely requirement for high gas rewards, which gets us back to time value predictability).
Some things to consider, when trying to estimate the future potential of gas rewards:
1. Currently Fusion is doing around 5 txs/block.
2. Gas cost/tx is to a large extent determined by supply and demand.
3. Early stress tests of Fusion gave it a claimed capacity of 2500 tx/second
4. Critics such as @Iruwen claims true real world capacity is more close to 25 tx/second (which is about 300-350 tx/block)
Out of 3 and 4 I have no idea which is more true, but I believe it's safe to say that if current network use (1) goes up 60x (so we have a situation of around 300 tx/block), we will to an extent begin to see the effects of 2, which will cause gas price to increase for regular txs. And that is probably what we need if the gas use/block is to become significant. However, there is also another factor that can help and this is that certain special txs, have a higher gas use.
-An extra fee of 0.001 FSN for Quantum Swaps and TL transactions.
-An extra fee of 0.1 FSN for USAN generation and asset creation.
Out of these it's most likely that QS and TL txs eventually will play a major role, and if there can be 100 of these tx-types per block, that already amounts to more than 1 FSN in a gas block reward.
Only once we have an active Swap Market compatible with FRC20s that is also properly advertised, can we begin to see whether such a tx frequency is likely to occur at any point. @djqian hopes that Chainge can bring about all of this at once.
What happens if it doesn't?
Things that are happening now, are likely primarily important for the coming 2-3 years. First halving is coming by the end of the summer 2021, which is when block rewards go from being 2.5 FSN to merely 1.25 FSN. But since this is still quite a decent reward (interest yield 10-15%), I believe gas use only truly becomes a critical factor at second halving (when block rewards become a mere 0.625 FSN, bringing likely interest yield due to the set block reward down to 4-7%), which will happen in the late summer of 2023. I believe it's quite likely that a good balance may be found by then, as long as the network is still being used. Because regardless of configuration, network use is what brings value, but making sure there will be a perfect balance between validator insentives, interest extraction and network use costs is also something that may help make the network popular. A good balance here is probably more important for success than network efficiency. If Ethereum has taught us anything, it's probably this.