Windows Core voices with NVDA now speaking the 'A'

In our documentation on using NVDA, we recommend not to use the Core Voices. I no longer see this as a problem. In Word with NVDA 2025.2 beta 3, it works in Word and When MathCAT reads the expression. I am going to remove this recommendation on the NVDA documentation I am working on.

Start by locating the NVDA documentation section that recommends avoiding Core Voices, then review the surrounding guidance and the reported Word and MathCAT behavior. Done means the outdated recommendation is removed and the documentation accurately reflects the current behavior described in the issue.

Written by the indexing model from the issue text.