Remember when games could be made in less than a year for $200,000? Or in two or three years for $20,000,000? Pepperidge Farm remembers.

Yep, even in the PS3/360 days we got whole trilogies in the one generation like Mass Effect, Bioshock, Uncharted, etc.

That gen we got 4 games each from Rockstar and Naughty Dog, now it's the 6th year of the current gen and neither have released a new game yet.

I was expecting the Hitman flavor to the game but I was pleasantly surprised by the combat which I agree is very reminiscent of Uncharted, there's also a bit of the Arkham games in the hand-to-hand combat. All together a great mix that works surprisingly well.

I don't know if I would call this the norm. If it's the norm then it's down to poor management and priorities.

Capcom, FromSoftware, Insomniac, and Monolith are very productive. Capcom is way better now than they were in the PS360 days, are they even the same company?

Square's Hamaguchi team is only taking 3 years to develop a new Final Fantasy 7 "remake" with large scale, tons of content, and excellent production values. Not to mention Revelation is releasing on all platforms simultaneously.

Inefficient AAA developers need to take notes.

There are thankfully still some devs who can work efficiently, yeah. It's just the frequent 6+ year cycles that are a pain in the arse.

At any rate, the game has apparently moved 2.7 million in it's first week: https://www.eurogamer.net/io-interactives-james-bond-game-007-first-light-sells-27-million-copies-in-its-first-week-as-dev-reiterates-free-year-one-content

Hopefully it's able to recoup its costs.

Some important context though.

FFVII remake (2020) was in development for around 6 years (announced 2015)  and restarted/shifted team at some point in late 2016/early2017. It also uses UE4 and doesn't really make use of any unique/bespoke technology like 007 does. It actually is very comparable to the development 007 whilst being a smaller scope game in some ways.

Only after this did they manage to produce Rebirth so smoothly (although it has some technical weak points). The direct sequel to 007 should come sooner than rather than later provided the new publisher doesn't mess things around lol

Doing new things takes time and a bit of trial and error. 007 is a whole new scope of  AAA game for the team so it's Developement cycle taking time makes sense. All the meanwhile they released hitman 3 in 2020, the collection in 2022 and are working on a new IP, so  the team haven't been exclusively on it for the last 6 years.

Insomniac and Capcom do show development can be insanely efficient in modern era when managed well. The whole industry ought to the learn for them, I wholly agree. I'll just add that their huge size probably makes the Dev cycles look shorter than they are. After all wolverine was announced 5 years ago but the team has released many games between  2020 and now, so they're really good at managing multiple titles. Probably their games average a 3 year dev cycle of individual developement. 

It is still possible as long as you don't spend forever on design/graphics/story lines.

If you "keep it simple stupid" (kiss) and don't aim for the moon, you can still make games without it costing a farm.

The problem is then your in the "indie" space and basically forced into Pixel art territory.
Unless you "cheat" like 2D backgrounds with 3D character like in old PS2 games, or such. 

However if you just super cut everything out of the studio (HR, secretaries, sweet baby consultant... absolutely everything not a coder basically...)
You can still pull off miracles like Clair Obscur: Expedition 33.

Remember that had a cost of less than $10million to develop in total (8.8m + 1m for voice/animation of characters ect).
This took like 5years with scaling up, where early on, it was just a few people bouncing idea's around, working on the story.

If they knew in advance were they were going, and scaled up sooner, the development timeline could probably have been shorter.
They had to basically teach themselves how to do certain jobs needed, find random people to fill roles themselves.. 

They lost the opportunity to announce 0.07M sales in the first 7 minutes available

But the budget being $200 million suggests it's been in full development for a long time. For a short game, it's just not a good look. But this won't matter if it keeps selling well and spawns successful sequels that take 2-4 years to develop.

Like you said FF7 Remake development semi-restarted in 2017, to be built in house with no more involvement from CyberConnect2. Though it's safe to say they reused some assets from the previous work.

Rebirth and Revelation use assets and systems from previous games for efficiency, but Rebirth had a massive scale and so much content. It's honestly incredible what Hamaguchi's team achieved for a budget likely half or less as big as games like 007 First Light or TLoU Part 2 or God of War Ragnarok.

That's why I've said that games need to stop being so damn ambitious. The problem with games isn't that the tech keeps getting more advanced. It's that video games are tech. And the big publishers feel this need to keep one-upping themselves and each other, making the biggest most grand cinematic experiences they can with with increasing computing power, storage space, etc., they have at their disposal. It's like they feel they can't not use the increasingly powerful hardware at their disposal. The problem is that we are firmly in the area of diminishing returns. Video games are getting close to looking as good as they're ever going to look. They also don't need to keep growing in scale and scope. If 40 hours isn't long enough for someone, well, I don't know what to tell them.

In any case, this is the only entertainment medium where we've seen this massive amount of growth in scope and scale. In 40 years we've gone from games like Super Mario Bros. that you could beat in a single afternoon to games that can take a month's worth of sessions to beat, all while the visuals and presentation have gotten ever more sophisticated. Movies still average around two hours, and while special effects have become more advanced over time for the same reasons game graphics have improved, not every movie is a CGI-laden action-adventure film. There's a wide variety of genres and budgets, of use of special effects, of the special effects techniques used, of cast size, and so on. Not every movie has a $200M+ budget, and average budgets in general haven't grown at the exponential rate video game budgets have. Music releases are still the same scale and scope as always. Songs in popular genres have always been just a few minutes long on average, and are still typically released as part of an album. The length of an album was determined by the medium as LPs, CDs, and cassettes had differing capacities, but most single full-length albums run 40-60 minutes. If films or music grew in scale and scope the way video games have, we'd probably have movies or albums that ran for 80-hours, but I don't think anyone would want that. People expect a movie to last 90 to 180 minutes, and an album to last an hour or so at most. Also, those industries wouldn't dare do it because of the increased overhead costs that would come with making movies or albums that long.

Now, there's nothing inherently wrong with big, epic games. JRPGs were the equivalent of those sorts of games 30+ years ago. I could blow through most SNES & Genesis games in a single afternoon, but JRPGs took a bit more of a time commitment, with many of them taking 20-40 hours to beat. But they were the exception, not the rule. Now big games are the rule, and smaller ones the exception. But not every game needs to be some 40-hour epic. Publishers (and gamers, too, to be fair) need to stop treating "short" and "linear" as dirty words. There's nothing wrong with a single-player game with a solid 8 to 12 hour play time, either. Capcom is one of the few that still consistently nails this with Resident Evil and even some newer games like Pragmata. They even brought back Mega Man with 2D gameplay & 3D graphics (that aren't super-sophisticated & ultra-detailed) with Mega Man 11, and are now making Mega Man 12. I can beat Mega Man 11 in less than 2 hours on normal difficulty. Most of my favorite games are games I can beat in one to three sittings.

Also, I think the increasing ambition is probably in part something more personal for people in the industry. Video games are still not taken as serious art by many in other fields in the arts & entertainment, and a lot of people want to prove games can be art. In the process, the AAA segment has created what are essentially 40-hour interactive movies. But gigantic open worlds and hours of mo-cap cutscenes don't come cheap. My response to "Are video games art?" is "Who cares?" They're interactive entertainment. They're their own medium. Let games be games. They're supposed to be fun, and that fun can come in different forms. Some people like a story-driven epic RPG, some people like a platformer with a token story (like every Mario game ever), some people just like competing against others in a multiplayer experience. Video games have the potential to be everything from a simple pixel art game that can be beaten in an hour to some juggernaut like Grand Theft Auto to some couch co-op party game to a game where people can just goof around and build things. Big publishers need to diversify and stop dumping everything into bloated 60-hour open worlds or some live-service title that needs constant updates & maintenance and will probably struggle in a market that's already oversaturated. And they need to figure out how to cut overhead without resorting to overworking employees or making them use garbage like generative "A.I." (or Plagiarism Bot 3000, as I like to call it).