Beer recipes are everywhere. Every jackass on the internet posts recipes, it seems like. Step one is sorting the wheat from the chaff, so to speak, and we all have our own tricks and tolerances there. Let's assume, though, that you've found a reputable recipe: it's still going to need to be adjusted to fit your system and process. What's the best approach for doing so?
The Procrustean Method
Procrustes is a character from Greek mythology, who had a bit of a sadistic streak. He lived along the road to Athens, and he'd invite travelers along the road to spend the night at his place. Once there, though, you had to sleep in his special iron bed. If you were too short, he'd stretch you to fit; too tall, and, well, you can probably guess.
Lots of people seem to take this approach to their "sourced" recipes. They adjust every recipe the same way, fitting it to their own Procrustean Brewery. That's not necessarily a bad thing, and it'd be out of character for me to argue against a systematic approach to something, but I'm still (kind of) going to do it.
First, the good: it's a smart idea to make identical adjustments to correct for idiosyncratic problems that are reliably caused by your equipment, or process, or ingredient supplies. If you know you have hard water, or low efficiency, or you bag your hops, then sure, make those Procrustean adjustments.
However, don't become a slave to those adjustments at the expense of looking at the what and why of that recipe you're adopting and adapting. Most of the recipes you'll want to take on will probably be for beers with unique or distinctive characteristics, and if you fixate too much on the mechanics there's a risk that you cut corners elsewhere and/or don't fully commit to the recipe. Which brings us to...
The Lebowskian Method
The Dude is a character from the Coen Brothers mythology. He lived in an apartment, and one night some visitors rough him up and defile a rug that really tied the room together, man. Hijinks ensue.
Lebowski (the Dude) is, arguably, a drunk, stoned moron. But he teaches us an important lesson: keep an eye out for the things in the recipe that really make it what it is. Which of those ingredients, elements, or steps really "tie the beer together?" Maybe it's a specific grain (Fawcett 45L British Crystal) or a process (decoction). If you're adjusting a recipe to your brewery, you might end up wasting your time if you substitute in your normal American Crystal 40, or reduce the amount of chocolate malt you use "because your roasty beers are always astringent so you always cut the recommended chocolate malts by half," or decide to just do your typical infusion mash - and no amount of your Procrustean tinkering is going to fill the hole that's left in your beer.
Give Both Their Due
Good recipe adjustment requires a bit of both. I write a lot of recipes for brewing publications, and I'll share a not-a-secret secret: when I share my recipe, that's not my exact recipe. We use standardizing spreadsheets to ensure that each recipe has a common point of reference on efficiency, water-to-grist ratio, hop utilization, etc. so that you can hammer away (like Procrustes) and make it fit your system. As a result, what you read is usually already different than what I actually do when I brew the beer. We expect that you can and will make changes to weights, volumes, times, and more. Procrustes should get to have a hand in your beer (so to speak).
What I do spend a fair amount of time advocating for, though, is for some specific ingredients of steps that make it much more likely that you'll get the "special" part of the recipe right. Even if your IBUs are off, or the water chemistry isn't quite on target, or you miss my OG by a few points, it'll matter a lot less than if you sub in that "house" malt or yeast of yours for one that the recipe calls for, or ferment it at 65F instead of 52F. Keep an eye out for what ties the beer together, and commit to matching that part of it. The Dude Imbibes.
A Simple Approach
When I get a recipe I want to adapt, I take a pretty simple approach. Each will be a little different, of course, but this might work for you as a rule of thumb.
1. Adjust for efficiency differences, but only with the base grains, to hit about the right OG.
2. Leave the weights of the specialty grains alone (unless you're doing a wholesale scale adjustment, like from 10 to 5 gallons).
3. Use exactly the ingredients listed for the grist, and any hops added within 30 minutes of the boil (early hops may add a very slight detectable flavor, but mostly just add IBUs, so substitutions probably won't hurt you).
4. Use the recommended yeast unless you can't get it fresh, and even then, pick the nearest substitute (so, London Ale for London Ale III is fine, if the LAIII is two weeks old and the other is two months old).
As for process changes, go with your gut and let parsimony and commitment be your guide. Extended boil? Sure - no reason not to. Water adjustment? Maybe not - you might do more harm than good. The deciding factor should be whether you have good reason to believe that the recommendation is a key feature of the beer's flavor profile, and if you decide against doing it you should also mentally commit to trying the same recipe again with the change in the event what you get without it just doesn't work.
So take it seriously - but also don't lose the forest for the trees.
Keep it simple.