Another thing is that there are plenty of factors that go into designing a GT. We can talk about Froome-proofing a GT, but I suspect we may see less of that going forward, for the simple reason that the fans' rejection of Froome this year was far less aggressive than in 2013 or 2015. Whether it be because we've just got used to him, or because he raced a more interesting race this year (and there was more frustration at others not challenging him rather than at him as a result) and there was no "Clinic explosion" dominant performance like Ax-3-Domaines or Pierre-St-Martin, or whatever the reason may be, one of the reasons for organizers trying to design routes to allow others to get close to Froome is quite possibly because they realized they have an unpopular champion; if somebody dominates and is popular, it can galvanize the sport's fanbase. But if somebody dominates and is not popular (notwithstanding that Froome's success being consistently followed by doping innuendo in the media and on social media) then fans want to cheer for somebody to dethrone them, or will just switch off in frustration. Obviously ASO don't want that. If the fanbase is ready to accept or at least tolerate Froome as a major champion, they won't need to try to engineer a course to prevent him winning too easily anymore.
At the same time, organizers have done plenty of tailoring of courses towards the riders they want to win. Back in the 50s and 60s especially the Vuelta used to tailor its course directly to the foreign stars it wanted to turn up, in an attempt to secure their participation, even at the expense of their own homegrown stars. This only really started to change in the era of Ocaña and Fuente (although the race director did in fact ask KAS' DS to call Fuente back in his legendary Formigal raid to prevent the GC being settled with a comparative unknown winning, with the KAS DS understandably and thankfully telling him to go forth and multiply). We still see it to this day though; the way the Tour removed the TTT after 2005 and then didn't have one for three years, but when Armstrong was returning we saw a long, difficult TTT early and then the first set of mountains being neutered, it was clear a lot of the route was designed around the thinking that, if Lance wasn't all that good (we had no idea how strong he'd be in his comeback at this point) then he'd at least be in contention for the first two weeks, guaranteeing the spike in public interest. 2012 is a similar example, where it seemed fairly transparent that the organization wanted Wiggins to win and capitalize on the new-found British interest in the sport (coming off the back of a number of other ASO-organized races which favoured a strong time trialist who could control the race).
At the same time, we've seen a parallel development in the Giro and Vuelta to horrendously imbalance the routes because of the audience figures, as previously mentioned. ASO will always have less trouble attracting the world's eyeballs at the Tour, because the Tour is the one race which has truly global currency even outside of the cycling fanbase. Zomegnan and Guillén, although their philosophies related to these were different, both noticed that the increased audience figures for hilly and mountainous stages were notable, and tailored their routes in that direction to capitalize. Zomegnan, of course, went a bit too far, and of course lost his job, but Guillén is still in his role. He got very lucky in 2012 that both the Giro and Tour were absolutely awful that year, and that a range of lucky circumstances led to his route, an experimental affair with no stages in the south, some absurd transfers, a dozen mountaintop finishes, next to no major mid-stage cols, hardly any time trial mileage and some absurd circuits, being a success. Even so, just like this year when the same formula was used, it was the least anticipated mountain stage that created the story that people will remember.