Arny Krueger wrote:

And isn't the
dye made of metal?

Dyes are complex chemicals made up out of lots of different elements, not
just metals.

I should have written "partially made of metal".

Nevermind, the point was the loss of data because of
recrystallization.

If you abuse a LP or a CD there will be data loss. What's your point?

In this case, there's no "abuse". The recrystallization is a natural
tendency.

The description made was certainly inprecise and
inadequate, but the phenomena exists. Ask a physicist.

The question is not whether the effect exists, but whether it's a problem,
practically speaking.

Most people will certainly think it's not a problem, because the process
can take between 15 and 50 years, perhaps more in very good storage
conditions. It depends also of the composition of the dye. I assume that
a very close examination of the very first CDs made in the 80's (so 20
years ago) would give an idea of the time it will take before the CD
becomes unreadable.