If the files were used in a game then you have corruption, they are not though so you do not really have that*. If the files were there but of unknown type and you could repoint to them, you tend to see this more with music (ten audio tracks detected, only 8 used in the game, tell the game to instead play tracks 9 and 10 via simple pointer changing, maybe bank changing as well and the game will hopefully play the tracks without you so much as knowing word one about the audio system) but the principle works for most things with beta/test levels, 3d models and such are a distant second for that sort of thing.
File system using consoles are great so you can open up a game, look at things and try to figure some stuff out. However once that fails you will rapidly run into needing hacking skills -- hardware knowledge, format knowledge (existing and general concepts that underpin it), possibly some assembly knowledge for the device you are on and more besides.
*if you open the game, see 10 files of a given type (size, extension, name, folder they are in, header, general look of the hex...), corrupt them and things change but 2 do not appear to do anything in the game you might well have some unused/cut content and can use that to direct further fiddling.
"There was one image present that was posted... Unless the people who wrote the articles both lied, the data must be out there and be obtainable. "
Though lies are not unheard of you also have demo versions (quickly slap together a demo before you go gold but are otherwise in crunch time or you get Johnny the intern to do it....), developer/early reviewer versions (if IGN or someone got sent a game back in the PS1 era it might well have been many months before and played on almost dev hardware), PR images from the dev/publisher themselves (compare e3 to the final released version...), actual beta versions that might have leaked as well as the region/earlier pressing thing already mentioned.
As for the arcade version then much the same way as you approach any new systemhttps://github.com/mamedev/mame/blob/704594332184de279135d29028df9dae0894acf6/src/mame/drivers/namcos12.c
aka Namco System 12
and a whole bunch of other useful data, not so much for ROM hacking directly (it is not going to tell you what does what but it will cover various higher level memory mappings).http://www.psxdev.net/forum/viewtopic.php?f=49&t=65
links to a bunch of things on the CPU, a nice little instruction set summary to begin with ( http://vhouten.home.xs4all.nl/mipsel/r3000-isa.html
) and some encodings in the next one ( http://www.mrc.uidaho.edu/mrc/people/jff/digital/MIPSir.html
) and later links ( https://web.archive.org/web/20061010210946/http://decstation.unix-ag.org/docs/ic_docs/3715.pdf
) getting quite fun indeed.
If you have the choice between learning this and the PS1 I would probably learn the PS1 first, the hardware is not crazy different though.