I know this is a type of boss design (most commonly found in Japanese RPGs) but it is one I personally hate. It is the polar opposite design philosophy of what Metal Gear used to be, which was more puzzle-oriented, like Zelda. Metal Gear bosses used to be about learning patterns, exploiting weaknesses with specific weapons, crippling the enemy to give yourself an advantage, etc. The bosses in Peace Walker swing completely in the opposite direction, into stat-driven endurance battles. This is where the Monster Hunter influence goes too far, reducing Metal Gear to a straight-forward grind-fest.
I love the Pokemon stuff, the kidnapping and army building, but in some ways it was better in Peace Walker's predecessor, Portable Ops, when these elements were simply a meta-game laid over a core game that was still recognizably derived from classic Metal Gear. While it's true that Portable Ops marked the first time bosses lost some of their puzzleiness (mostly as the result of letting players design their own arsenal) they never required grinding to win.
Unlike in Peace Walker, weapon and tool development in Portable Ops was holistic, not incremental. In other words, items did not have various "levels" of power or effectiveness. You didn't have to "upgrade" your rocket launcher to make it do more damage. A rocket launcher was a rocket launcher, and you either had one or you didn't. Sure, there were the RPG-ish elements of needing scientists to build weapons, and what they created and how fast they created it were based on a rudimentary stat system, but once you had an item in the field stats didn't matter. It was about which weapons/tools you had, not what "level" they were.
I can't stand the way Peace Walker scales difficulty by scaling enemy statistics. This essentially means the only way you progress in the game is by scaling your own statistics. It's less about how good you are and more about how many fucking rations and supply markers you have, so you reach a point where you simply outlast the enemy simply because you put endless hours in the game. It's the kind of game design that devalues learning and skill in favor of not having a life.
If there was any doubt about Peace Walker's "damage sponge" difficulty philosophy it is proven by how the game omits any and all permanent effects that might give players a strategic upper hand. Setting anti-tank mines or blowing up a fuel tank only stops land vehicles "temporarily" even though they should in all rights stop them permanently. It's clear each boss is designed not to be "too easy" for players who want to pound away on it with their snazzy guns. Since everything has hit points now it's just a matter of hitting bosses--anywhere--until they go down. This is a far cry from the tank battle in Metal Gear Solid 1, where one grenade would disable its treads and another down the top hatch would finish off the gunner. The main challenge was getting close enough to the tank to do this, and the fight was perfectly interesting, logical, and satisfying.
Given how excellent the simple puzzle-logic of Metal Gear boss fight have been in the past, it feels dumb for Peace Walker to simply abandon all of it in favor of straight-up RPG stat-grinding. The better fights in the game--the PUPA, the ZEKE fight, and if you choose to try and stealth the vehicle bosses--retain some of the old Metal Gear strategic thinking. When it comes to the later bosses, though, it's so stat-heavy and grind-necessitating the game feels more like Dragon Quest than tactical espionage.
I always loved Metal Gear's reliance on tools with discrete uses rather than stats with incremental effects. This is what put the series in the same category as Thief and Hitman--all superb games about using sharply-defined tools to make decisions in a richly simulated world. Peace
This isn't to say stats always work against strategic decision-making. It depends on how they are implemented. When they seem to exist only to augment things like health or damage they do. But when used in other ways they don't. Metal Gear Ac!d, the short-live Metal Gear spin-off series released on the PSP some years ago, indulged RPG conventions without undercutting this sort of tool-decision-making. It's hard to imagine anything more RPG-ish than Ac!d's turn-based, card-based combat system. Yet I have to confess that--when put side-by-side with Peace Walker--both Ac!d games manage to express the strategic thinking of classic Metal Gear in a way Peace Walker seems to totally lose sight of.
Even though Ac!d featured a "card deck", in which actions could only be "played" based on which cards happened to come up in your "hand", all these actions had discrete functional values, not arbitrary incremental values. Drawing the card of a particular tool or weapon meant you got to use that particular tool or weapon. Pistols, rocket launchers, etc. all had specific strategic values. It wasn't just about how powerful they were. There was no rocket launcher "+1" or "+2" because challenges did not scale primarily in terms of how much HP enemies had (like they do in Peace Walker). Like any true turn-based strategy game, the Ac!d series was all about, well, strategy. It was about how well you could out-think your opponent by seeing several moves ahead of them and using your resources accordingly.
I remember spending hours on some screens of Ac!d, just trying to figure them out like puzzles. I specifically remember a screen full of snipers perched on ridges, and having to figure out how to use my current card deck to sneak past them. It was hard but rewarding once I developed a successful strategy, the way any turn-based strategy game is. In this sense Metal Gear Ac!d recalled Front Mission, Vandal Hearts, or even the original X-Com--all turn-based strategy games where cleverness was more important than how high you had grinded your characters.
Metal Gear Ac!d was a PSP launch game, and at the time I remember Hideo Kojima claiming he was skeptical as to whether the real-time tactical stealth gameplay of Metal Gear would "work" on a portable platform, hence Ac!d's "experimental" turn-based approach. Ac!d was predictably criticized at the time for "not being a real Metal Gear game" even though most people admitted it was quite good turn-based strategy game. Portable Ops, in obvious response to this, was intended as the the first "real" Metal Gear game on the PSP console, and Peace Walker was even more hyped as a full-blown main series installment, even though in some ways Ac!d was more true to the concept of tactical espionage action.
Thinking about Ac!d again makes me wonder if Peace Walker's more frustrating battles would actually be fun if they were turn-based. Even if they were they probably wouldn't be as fun as Ac!d, because they'd still be just endurance tests, which is the least interesting type of strategic problem I can imagine. Two opponents hit each other until one of them dies. Brilliant. If I wanted that I'd play...
...well I wouldn't play Metal Gear, that's for sure.