For me, the reasons why I think Gold for Exp is so alluring has more to do with the knock on effects of making attaining treasure the primary focus of play, for context the version I'm looking at for our Pathfinder 2e West Marches would involve paying to level, rather than just getting exp when you acquire gold, or when you spend it on whatever.
1. Treasure is an amoral motivation-- whether the characters in your party are good or evil or neutral the need for funds, wealth, and magic items is a believable motivator. This means that characters of differing moralities have a common goal for their adventuring, and that as a GM I don't need to worry about or police anyone's motivation to adventure. My variant capitalizes on this because since the treasure spent on leveling is only used for that purpose (as opposed to being able to pick up a magic item at the same time) the flavor of what makes you level can be bent to support your character's morality: the default might be paying someone to actually train you, but its acceptable to say you put the wealth toward an orphanage, or sent it home to your starving village, or used it to fund the operations of a cult to your evil god, or donate it to the descendants of the people who built the ancient ruins you just plumbed, or effectively donate your cut of the stuff to a museum. Similarly, secondary objectives, like rescuing hostages can still come up, but they aren't the sole motivator, meaning characters who aren't interested can always fall back on the possibility of finding treasure anyway or the need to preserve the party as a source of future wealth or some such, but if your character WOULD care about that, you still can.
2. Treasure works great for nonlinear content because it represents a granular and an expansive nonbinary victory condition, if you have a dungeon where treasure is the goal, then all paths that offer the possibility of treasure are viable routes (as opposed to say, the single path that leads to a boss monster that must be defeated for the good of the realm, that make all the other paths only obstacles and distractions) and the party can experience partial success by getting part of the total wealth stashed there. This inherently supports interesting adventuring locations that don't have to be straightforward, and where the party can focus on exploring the space without developing tunnel vision on a single point or a solitary objective. This also supports sandbox play by allowing the party to choose how much of a dungeon to explore (rather than a 'but thou must!' due to the story consequences of the location demanding they address some particular threat.)
3. All of the above can have the additional effect of refocusing the game on personal stories by providing a reason to adventure that doesn't have to hog the narrative spotlight. I like to call this 'adventurer slice of life' where the goals of the adventure take a back seat to the relationships and personalities of the characters, along with whatever situations they find themselves faced with. More traditional goals tend to make the adventure about the destination (beat the BBEG!), rather than the journey. Similarly it can make downtime more believable where the characters settle to enjoy their successes before planning a new expedition, which is nice because it helps avert the "zero to godlike in a few weeks" plot holes that seem to riddle a lot of games.
4. In my game, using wealth to level helps the West Marches format by empowering players to set their own progression speed-- they can ensure they're well kitted before moving on, or they can rush to catch up in level with their friends who play more, they can weigh their preparedness against the possibility of taking on higher level (and therefore more rewarding) content. It also helps the hexcrawling because getting the treasure back to a friendly port is necessary to 'finalize' your acquisition of wealth so all kinds of pirate like events are possible, like ambushes by rival crews. Similarly we have rules about how players are going to gather parties and schedule sessions with GMs, and be the ones to figure out their own cuts, which has cool implications since player owned ships, and paid hireling crews are a part of this too-- the GM is just going to enforce whatever distribution the players agreed to prior to setting out.
5. It can encourage information gathering proactively, since the players can effectively always be looking for new sources of wealth, I love stuff like this because it really makes all those cool little simulation elements in systems actually get used-- things like 'Gather Information' or 'Research' in Pathfinder 2e, the lack of a plot means that there's nothing to grind the train to a halt if the players don't know what to do, they can always just search for new leads on where to get treasure, there's always an answer to 'what do we need to do next' that isn't "look for monsters to kill" or "give the GM puppy eyes for the next bit of plot."
6. If you have to actually spend money to level, it makes leveling as fungible as your gold is-- if you wanna prop up a buddy or a secondary character or something, you can literally pass wealth around to facilitate that, which in a leveled West Marches, is nice-- you can invest in your fellow players, or in having more options for future expeditions.
7. It provides a natural tradeoff to the decision of how many players are even invited to come along on any given adventure-- we're scaling the adventure for four pc's of the level of the lead in terms of both encounter guidelines (though GMs aren't restricted to balanced monsters, not everything is meant to be fought), and treasure, so bringing more people is intrinsically safer... but probably means they're going to want more of a cut. Because gold scales exponentially as you go up in level, there's a significant pay off to trying to punch a little above your weight class, but that is very much something that will be very risky, and demand touch and go tactics and weighing how far you can push your luck.