DNA is expensive to maintain, from a biological stand point. Why would nature select for simple pairs chromosomes but fill them with 90%+ junk. Why not "spend" your DNA budget on tetra+ chromosome groups and less junk?
Because such clean-up requires a mechanism, which doesn't exist until it gets evolved. Carrying along some broken-and-disabled junk in long-term program storage doesn't cost enough to be a strong selection pressure, so under just mutate-and-select it can hang around for a long time.
Also, if it is the remnant of a functional subsystem, until it drifts too far it might again come in useful, randomly get turned on again, and provide an advantage. After it drifts into doing something somewhat different, it might get turned on again, prove adventitious, and become the basis of a new system.
Lifeforms don't have to be built to superb standards. They only have to be good enough to work, and to work well enough to not get squeezed out by something that works better.
So it's entirely plausible that a lot of DNA junk might be present.
Which is not to say that there ISN'T one or more DNA-cleanup mechanisms currently in place in some branches of life. (Probably evolved from old retroviruses, like several known mechanisms for DNA alteration.)
But it's a hard problem. (How does the mechanism know not to take out something necessary? If there's something it avoids chopping out, what's to keep junk DNA from masquerading as good DNA to hide from it? Junk DNA subsequences are still subject to evolution and hiding from a DNA cleanup mechanism is advantageous to the DNA.)