In case you were unaware of it, or indeed some hacker wants it to make life easier, then there is a save with the gamecube content unlocked. For those thinking "ew GBA film tie in" then while 99% of the time you are correct this franchise on the GBA did have something to it.https://gbatemp.net/download/lord-of-the-rings.27187/https://gbatemp.net/threads/gba-return-of-the-king-sav-file-gamecube-content-unlocked.320305/
There is also a trainerhttp://www.advanscene.com/html/Releases/dbreleases.php?id=1219
As far as the hacks in question.
Increasing save slots is possible but rather more than I would suggest for an earlier hack (Eeprom v122 - 4 kbit, if ADVANsCEne is accurate, is about as small as you get on the GBA -- flash several megabits technically being available if people are concerned about overall size). Save type conversion itself is trivial (the save patchers/SRAM patchers of flash cart patching software is literally this, think there are even some open source implementations these days both generally and for flash carts) but getting the game to do much more with it/handle things is the trickier aspect. I would almost be tempted to run two versions of the ROM with separate saves on my flash cart and go with that instead, porting around data can also be done reasonably well if it comes down to it and you want a specific team (chances are said character is a section in memory, slice up a savestate or make enough cheats to port it over and you have that done, if you want to hack the base save, though see save hashing for that one, then you have an option there as well).
Same for general concerns inventory slots unless it is a game wherein you unlock more as the game progresses. Possible alternative for inventory slots is some kind of cheat that cycles things in and out*, or maybe a pointer tweak. More generally then if you have 8 inventory slots then you probably also know how much memory that will take (you in this case being the hypothetical game coder or compiler) and thus rather than waste memory (you have a little less than 300 kilobytes to play with http://problemkaputt.de/gbatek.htm#gbatechnicaldata
) will slam the next section of data right where the other one finishes and so on and so on. Expanding one without changing where the game looks for all the others is very tedious, changing where one lands to a free section is easier (find it in RAM with a cheat, chances are there will be a pointer you can get to and change) but you have to find that and also make the game's save handle it (it might do it by default, it might not). You will probably also have to handle the game logic and UI level stuff to make it go beyond the previously known areas to read, and handle what happens when you get to the end of section. Sometimes this can be easier if there is a game level limit/purpose to the limitations but no technical one but from what I can recall and what you describe I am not betting on that here.
*should not be too hard to find extra memory. Whether it gets saved with the main save is a different matter -- GBA was new enough that copy and paste section was an easier path for some, however if it is going to have been crammed into 4kbit (512bytes, as in less than a kilobyte) then I might expect the devs to have gone a bit more old school and only picked that which they truly needed. If you are playing with savestates as your method of saving (most emulators but only a handful of flash carts and hardware level replications are this) then no great issue as long as the cheats remain active.
The GBA has powerful enough cheat engines ( https://doc.kodewerx.org/hacking_gba.html
, emulators and modern flash carts or even GBAATM/GBAATM rebirth should be able to do the more exotic stuff) to do button activators to check a flag, to grab certain areas of memory, stash them somewhere, grab others, overwrite them, set a flag to reflect the change).
Commissions are also a dubious concept in most ROM hacking circles -- money has a habit of making lawyers perk up (this sort of thing is already skirting the edges of laws) and even if not then you face the beer money vs real project concern. Beer money is great, if you want to pay average hourly/project rates for even a basic coder (never mind in this case someone familiar with ARM assembly) then that tends to be a rather more pricey proposition. This is not the first time I have done something like this overview of necessary changes to accomplish those goals for this game though.