I appreciate the further details and the link that you shared. This is the type of math I was hunting for to see exactly how the net limit was being proposed. I’m spending Saturday doing some more research on this governance action so I’m including the link you shared in my reading.
I also found this detailed breakdown of treasury spend by Jaromir that I’m adding to my research list: We Must Spend Responsibly From The Cardano Treasury 
Good question. Whether a fixed value or a percentage, the amount will become smaller per year unless Cardano can finally figure out at scale how to attract more people to the ecosystem and get more on-chain transactions.
Perhaps the answer is a hybrid approach: a set percentage when the treasury income is positive with a negative year triggering the requirement for the establishment of a temporary fixed value that’s lifted once treasury income becomes positive again?
According to Jaromir’s link above, treasury income is already in decline (negative) so following this narrative, using a fixed figure for now may make the best sense, while switching over to a percentage of income when Cardano starts seeing a positive inflow will allow access to the benefits of percentage-based spend.
Why use a percentage?
I’m thinking about it from the POV of efficiency and automation: with a percentage in place the annual spending amount is pre-determined without the need for manual involvement (or potential bickering) other than over how the amount will be split up and spent. An attempt to diminish bureaucracy, if even by a fragment.
Once Cardano has positive inflows the % can be set at a size that allows room for effective spend on maintenance, growth and development while also ensuring the reserve remains at a healthy size.
So those are my thoughts on percentage over fixed treasury draws.
My other concern is the fixed amount being proposed for 2025 may be too low. I see that at this point a lot of SPOs appear to be opposed to this GA, I have to take some additional time to understand why that is. Maybe they feel that amount along with what Project Catalyst draws and the proposed Intersect MBO spending puts their compensation in jeopardy?