Add more implementation guidance for thin-film iridescence

[peterkutz]

Think-film iridescence can be difficult to implement, and the paper we reference can make it even more complex due to advanced color-related transformations. This task is to add more implementation guidance to the OpenPBR spec based on our own implementations of thin-film iridescence, which are summarized on Slack:

Thin-film GLSL implementation (from @portsmouth ):
https://academysoftwarefdn.slack.com/archives/C06365EAQMP/p1710080253470669

Thin-film iridescence in Eclair (from @peterkutz ):
https://academysoftwarefdn.slack.com/archives/C06365EAQMP/p1732332598108709?thread_ts=1710080253.470669&cid=C06365EAQMP

More thin-film implementation notes: Improved iridescence on metals (from @peterkutz ):
https://academysoftwarefdn.slack.com/archives/C06365EAQMP/p1733848688599819?thread_ts=1710080253.470669&cid=C06365EAQMP