I can add a tidbit about planning bloody long triangles...

I'm now using SkySight weather software. It has some very cool features that have been added. You are able to position turnpoints are arbitrary locations within SkySight and the software will predict your speed around the course as well as advise the optimum start time and thereby inferring from that, whether or not the course is even doable in the day's available weather window. When a day is in prospect, I'm testing triangles in different directions and different geometries to see which produces the best speed. Once I have a general plan then it comes down to fine tuning by massaging the positions of a remote start/finish and the three TPs to get an optimal result. Thereupon, the selected coordinates can be saved within SkySight to a turnpoint file and uploaded to the flight computer for declaration.

If there are several successive days that are in prospect, then I do this speed simulation for each day and pick the best day to fly as I don't generally have the energy to fly long triangles attempt on successive days; I need to pick the best day to make an attempt.

As good as the software is, it's not perfect. No weather tool is able to get over-development and thunderstorms forecasted accurately as to where exactly the blowups will happen. So it's always a bit of a craps shot. It is necessary that the storm index / OD potential not be too high or you are almost certain to be blocked at some point when flying a triangle to declared turnpoints. It is quite different than OLC flying where a storm in one direction just means you change course to fly in the best air. Without that option, declared triangles are surely the toughest to get done.