Currently the input fields in Categories section of the Options dialog in VE does rely on placeholders as holding information necessary to understand the interface and fulfill user tasks.
It is preferable to feature a <label> element for the upper input field (and maybe refine the second one), from both screen reader perspective and for users with visual disabilities. I've tried to circumvent and improve situation for latter group in T139034, but there is understandable critical feedback around T85853 and follow-up T145676 about placeholders fulfilling level AAA contrast compliance.
