I wish WOTC had put guidelines in for how to adjust CRs based on an adjustment to those options. They give us dials to turn to make the game our own, but then don't list the CR formula to turn with those other dials turning. But I don't think that means the CR system itself is flawed, nor the monsters - the baseline remains the same and functional, and as long as it remains consistent then adjustments to challenges can be made consistently too. If you learn how to adjust based on one or more of those dials being turned, the CR system continues to serve it's purpose. But like I said, I wish WOTC had put in more guidelines on how to make those adjustments. A lot of people use those options, and it would be helpful to know how to adjust CR based on each adjustment factor.
For me, the basic adjustment is pretty simple - don't adjust the CR level or anything, but dial up the damage output of the monsters. For example, give them +1 to hit and to their spell DC's, per 'tier', is a baseline for the bare minimum adjustment required for a group with PC's that have grown characters from scratch (i.e. know how to play them in real life), and/or have some magic items etc. Maybe also pump up the attacks per round or damage per hit, but not necessarily. That will help avoid the common problems with higher level play, where the monsters just can't hit the PC's often enough, or if you dial up the challenge rating it becomes a huge slog where it takes hours to whittle down all the monster HP on the table.
BTW, 5e was extensively play tested at the very lowest of levels - all the open beta tests were starting at 1st level, and in my home game we also started at first and only got to about 4th or 5th level during the closed alpha and beta phases. The higher levels were only really "spot tested", i.e. people (including WotC people) made up a group of PC's at level X, and fought some monsters. That's not a criticism - they gave a long time for play testing, there's only so much you can do without giving it a couple more years of real-life play testing in big campaigns - it's merely an observation that supports the basic observation that the maths behind 5e only assumes PC's that are "plain vanilla", whereas if you run a regular game where people multi-class, use feats, optimise, know how to play their PC well, have magic items, only have a small number of fights per day, and so on... then yes, the maths doesn't quite work and the guidelines in the DMG / MM need adjusting as the group levels up.