feat(docs): reorganize

[jrasm91]

WIP docs reorganization efforts. Quick prototyping by telling AI to do a bunch of stuff, so will need a decent of manual review and tuning.

[immich-push-o-matic]

📖 Documentation deployed to docs.pr-30460.preview.immich.app

[multi-suggester]

possible changes i recommend for consideration or as a food for thought

• condensing some of the pages like facial recognition might be good as many different pages with small changes creates the necessity of extra reading (in concepts , in administation as improve clusters and also in faq atm)

• provide a references for facial recognition at concepts to faq section , hardware acceleration and improve facial recognition clusters (perhaps reference hardware acceleration to other pages like clip models , and jobs section as well?)

• consider reordering sub topics from name order to custom ordering that is either sorted based on commonness to nicheness or based on tech savviness? (or usefulness?)

• also i may recommend introducing more sub section levels as some parts like administration has gotten bigger/longer (like administration/customization(or something like external changes?)/remote machine learning sorta?)

• consider seperating some of the topics like user level , administer level (sub topics internal infos and customization(and settings) ) and external administer level (for things like remote access or scaling or external libraries?)(things that cant be changed from inside of immich and require outside changes and stuffs)

• (i wanna recommend moving some of the infos as sub sections of administration to keep the related infos in similiar places but i have a feeling you guys mostly want to keep it as settings sorta only)

• consider changing the text size of sub topics as currently menu on the left is bit long for scrolling imo (e.g. administration page)

• i also may or may not recommend adding a new section like "setting up" but with more customization themed (like install , hardware acceleration , remote access , model selection , mobile app/backup , oauth , storage template and scaling/remote ml stuffs?)

(lemme know if these are useful )(or not and ill avoid further comments :P)

[multi-suggester]

a detailed explanation about when manual tag effects the facial recognition results and when it doesnt might be a nice addition? (i roughly remember that ususally it doesnt effects but if face was detected by ml too then its added sorta thing? , not too sure)
(currently manuel tag isnt mentioned at all or i missed it)