Building/tuning is often just a lot of work. E.g. I recently tuned the Ford Roadster to get #1 on the rivals leaderboard for Fortune's Descent B class. I had to try every single type of tyre to establish that rally tyres would work best for that track with that car at that PI. Stock tyres were only a tenth or two slower. Race tyres and sport tyres were more like a second slower.
For upgrades, if you really want to optimise it, you have to go through them all and note down the change in PI and the change in power and weight, so you can combine the best value upgrades that will hit the target PI while giving the best power and lowest weight combination. For some tracks, though, you might get a faster time with more power and more weight, for other tracks it might be faster to have less power and less weight. When it comes to trying to shave a tenth off rivals times, finding a different combination of upgrades that can add 1 or 2hp of power or save 1 or 2lb of weight can take that time off. For things like brakes, gearbox etc, you have to ask if the PI cost will be worth it, e.g. for cross country I often don't upgrade the brakes, because braking distance just isn't very important, and extra weight can help you to smash through obstacles. A gearbox upgrade can be more or less free for some cars, but very expensive for others.
Johnson gives away a lot of good information in this video:
It's not all perfect, but it gives you an idea of some settings that many people might otherwise not try, e.g. the 1/65 ARB settings that all top tuners are aware of (but they don't work for every car, it's just a way to get the most out of the physics and PI for many cars by upgrading the rear tyre width, which is cheap, then using the ARBs to transfer the grip to the front, as front tyre width upgrades cost PI).