2.) Subjective question: is GLH worth considering if you have lots of coast on the same landmass (e.g. Pangaea with Natural shoreline), or is its main draw really the offshore islands / overseas land masses?
IMO no. This in keeping with the idea of a typical game mindset (meta settings and scripts) where the idea is to try to control as much inland area as possible and treat coastal locations as second choices.
GLH is one of the best wonders, if not THE best, as it is very early and almost always results in a net gain financially -- with enough cities that benefit from it. As trade routes are doubled with intercontinental cities, they are essentially worth roughly double same-continent coastal cities. The issue arises in just that though -- how many GLH-route-eligible cities are you going to settle on a Pangaea coast? 1/3rd of your empire? half? 3/4ths of it? Is it enough to offset their net cost? A typical game for me I end up at around an average of 12 cities (10-15 depending on map) before I call it quits on REX, as I love to play expand heavy. An empire of that size on immortal would need to have more than half the cities coastal to not tank with GLH route help, and it would STILL run at a net loss -- it just wouldn't be crashing.
Compare that to island cities, which not only break even (at least) with GLH routes, but they very often result in a net profit per city for commerce/expenses. They are always worth settling. In a similarly large empire as the previous example, any number of island cities would benefit me, and would only continue to as I added them, making the question of the ratio of coastal/noncoastal cities moot. Even just 2 island GLH cities can make the wonder worth it if you're staying smaller.
In that light I always evaluate my position on the map and weigh the 200h and tech detour against simply more expansion into the interior and faster BW/Pottery before considering GLH as a kneejerk gambit. If I'm on the coast and likely to stay heavily there (fenced in by a jungle belt/AIs, or many good seafood locations) I'll keep it in consideration until the turn benchmarks start coming up (want to shoot for 60-70 to get it done on Immortal to land it) and if there is too much else to do or deal with before snagging Sailing + Masonry I'll abort. Civ and leader picks can obviously tie into this just as much as the map can -- if you're on an AH only start as a non-AGRI/Hunting civ and are choked with forest, you don't have much incentive to go to the coast and tech in two directions at once, have you?
IIRC Natural Pangaea is the one with lots of little islands anyway right? The opposite of Solid? so it could be considerably more viable there, if only because it typically has 1-3 large regions of the map with tons of little islets sprinkled in very Archipelago-like chains, and if you start there you're money with the GLH. But just in a general land heavy script? I don't like it as much and would just ignore or go Mids instead.