> This is a widely held misconception. Mammalian genomes contain about 50% of their DNA as repeated sequences, and although there are people who can imagine uses for those sequences, there is no evidence that they are required or even evolutionarily advantageous.
If you remove noncoding "junk DNA", you alter the higher level spatial configuration of DNA. Histone winding changes which promoter regions are accessible, gene dosing is altered, binding affinity and kinetics change, etc. These are dynamical equations you're dramatically altering.
Non-coding DNA also likely shields against several classes of random point mutations, base substitutions, transposition, etc. preventing cancer and cell physiological disease states.
It's also important for maintaining alignment during crossover.
If it truly served no purpose, it would be gone.
> The idea that modern higher organisms have been "optimized" so that every part is essential
Essential is the wrong word here. Relying on every part for survival would put us at risk. We have plenty of built in redundancies to support degradation, failure, and loss of multiple systems and functions.
I get your argument, but I still disagree. At the species / population level, we have been optimized as wholesale organisms as best as development and body plans will allow. Everything not subject to pressure will get washed away.
Despite our vestigial tails, the coccyx supports our weight while we are seated. As I postulated before, the appendix probably has a net positive function in supporting gut microflora and our "junk DNA" plays a role at the molecular level.