Let me get one thing out of the way: The big problem with the 2017 live-action incarnation of Fullmetal Alchemist isn't that it's a production with a Japanese cast. The big problem is that a two-hour feature film just isn't the right form factor for this story, and that a Japanese feature budget just isn't enough to deliver the goods on such a story, not even in a cut-down form. This is a greatest-hits and highlights-from version of Fullmetal Alchemist, mounted and staged with the aesthetic appeal of a badly fitted toupée.
There was no good way to adapt Fullmetal Alchemist into any kind of two-hour movie, save maybe for making it into the first of multiple parts. That's what's been implied by this take on the material, which raids elements from the first couple of arcs of the manga and the original TV series, and cobbles together a distantly recognizable story from them. It opens with the young Edward and Alphonse Eric, honing their alchemy skills, and then attempting a forbidden process to raise their mother from the dead. It fails; the resulting backlash causes Ed to lose two limbs, and Al is only spared entirely from death when Ed fuses his soul to a suit of armor.
So far, so good, and the next sequence is also familiar: Ed and Al, how alchemists in the service of the Amestrian state, pursuing a religious charlatan who has been illegally using alchemy to hold his faithful flock in his spell. Word had it this fellow has a Philosopher's Stone, an artifact of such power that it might well be able to reverse the damage done to their bodies — and maybe even bring their mother back to life as well. But the stone in question turns out to be a fraud, and so Ed and Al have to resign themselves to investigating another alchemist whose experiments with chimera — alchemically created animal hybrids — may also have crossed a line. And meanwhile there's a trio of unnaturally powerful beings skulking around in the shadows and attempting to stop Ed and Al from snooping too close to something that could rock the state to its foundations.
My main objection to the compression job undertaken here is not merely that anyone remotely familiar with the material will feel like so much is missing. Adaptations have to be their own thing. I suspect the only reasonable way to make something with the scope of FMA happen with the opportunities and the budget available was to approach it as "part one of whatever" and cross one's fingers. That comes at a cost, though. Things that used to be dramatized and made emotionally concrete over the course of many episodes are now reduced to one-liners, or vanish entirely. A lot of what made FMA into FMA just doesn't make it to the screen, even by way of implication. You can't evoke that stuff by just nodding in its general direction. (In this version of the story, Ishval is barely even a name on a map.)
The real problem, though, is how everything else in the movie has also been drastically downsized and shoehorned to match. Despite some well-chosen location shooting in Italy, the end result looks like a big-budget movie that was red-penciled down to a fraction of its cost days before it went in front of the camera. Too many shots look suspiciously sparse, as if the effects team was supposed to add crowd doubling but didn't have the money. The interiors at Central look so underdecorated, I wondered if the prop handlers had gone on strike that week. And so on. When people talk about this FMA looking like a filmed cosplay skit, they're not just talking about the outfits.
It's the cost, not the cast, that was lacking here
The other thing they may be talking about is the elephant in the room that I've been tiptoeing around: the fact that this is a Japanese production with a Japanese cast, and how to Western eyes the original material lends itself to being something other than that.
I did my best, going in, to ignore how the casting might create aesthetic and cognitive dissonance. And for the most part, I found it all mattered a lot less than I thought it would. Most of the cast is more than up to the job: Ryōsuke Yamada makes for a credible, tightly wound Ed, and Yasuko Matsuyuki is excellent as the Homunculus appropriately named Lust; she could have vamped it up, but instead she holds back, and consequently it's more effective. I also liked Ryōta Satō, jolly and warm as Maes Hughes, who has all the moments with his family, and one involving a phone booth, that we expect from the character. (Veteran actor Jun Kunimura, a favorite of mine, also shows up as Dr. Marcoh.) Many of the individual moments with these characters work, if only as moments.
In fact, now that I think about it, the condensed story construction partly negates some of the issues that might have arisen with a Japanese-only cast. With the original story, you could make an argument that the Amestrians were Teutonic, that those from Ishvaal were Semitic, and so on, and you'd need to have casting that reflected those things. Because the film focuses entirely on Amestrians and a couple of Homunculi, the casting winds up becoming homogenous by default, and so any larger issues raised by it become all but moot.
But that just brings us back to the central issue with FMA'18. It's just too tiny a production, in the sense of both budget and story scope, to do more than hint at what its source material is all about. Granted, it was put together by people who cared about it and who know the territory. Director Fumihiko Sori previously adapted the manga Ping Pong to live-action, devised his own intriguing twist on the Zatōichi legend with Ichi, and gave us several original CGI anime creations (Vexille, TO, etc.). But all he's done here is reinforce my suspicion that sometimes being a fan of a given property makes you less equipped to do it justice, not more.