To this again btw, a big difference between leviathan and specific versions of the here proposed would be, that leviathan always is focused on high stat/level Illuvials and isn´t very flexible.
So unless you want to consider balance adjustments to force people to play/buy/level/farm different Illuvials than they currently use, you might have to "sabotage" the game itself to some extent to force different metas than reward efficient gameplay/comps.
Leviathan is more exploitable in that sense as well, that players can collude and type what comps they want to play, people do that in TFT as well, with just rating on the line, so no one else buys the same Illuvials, making needing everything less relevant unless it is very top heavy rewards wise. Top heavy certainly does increase the chance of exploits around matchfixing in various ways though compared to flatter rewards.
People can also in theory forever play with their Illuvials unless they aren´t allowed to be played/balance-wise unplayable.
If valid meta-strategies evolve around Tier 1-3 Illuvials (Stage 3s?), the demand for Tier 5 Illuvials, which if I remember correctly are the cost point of a full collection might not be increased a lot.
(obv some assumptions on if players behave more towards reward efficiency than organic gameplay if rewards and cost benefit calculations wouldn´t exist, but those are risk points, that could go wrong in especially the leviathan model)
Leviathan also likely can´t be balanced for the majority of players, as they don´t use the same assets, which should make it maybe not impossible, but harder to fine tune incentive systems.
Most versions of how to apply the here proposed would lift those potential (imo likely, but yeah assumption) issues, as at a new Set release you could shift rewards to new Illuvials, or during a Set you could have waves of rewards, making people not "get away" with budget rewards optimized options.
You can also focus game balance on purely game balance without rewards in the back of your mind, as maybe some balance could mess with the incentive system in a negative way.
This is less on the exploit side, but think I it is a good example of impossible to exploit in a sense of spending less than you earn, that´s doable with the proposed system.
Applying it to more than Illuvials would allow upfront cost+ gameplay options to get an item cheaper, similar to what battlepasses do, but for specific cosmetics people can choose.
e.g. new battleboard for 30$ every 2 month, daily top 4 during these 2 months with newest battleboard gives 0.20$ in rewards, if someone did top 4 every day give an extra big reward.
This could generate a relevant increase in DAU and would be a somewhat gamefied version of rewards in terms of daily quest kinda setup, compared to #leaderboard or placement based ones as well, which feel less like daily quests, more like bigger harder to obtain goals.
Now we´d have to make an educated guess on if and by how much this could increase DAU, as an increase in those makes most extraction/exploit options harder to pull off.
(To be fair to the example here a relevant variable would be how much opportunity cost is lost via these rewards, similar how it is with Illuvials)
In the end for both system to be as little exploitable as possible, the base the amount of rewards would have to be based on something like money spend to mint Illuvials or in Gauntlet (Cosmetics) in general.
The problem of Leviathan assets being likely usable for rewards for a long time period or forever even would persist though, as I understand that a new Set would likely not replace every Illuvial in Gauntlet e.g., while with other systems that could be avoided.
In the proposed System the risk of paying more rewards than people spend could at least be reduced to 0, which seems very hard in the Leviathan Model.