In bug 1745660 we discovered that, in some very rare cases (fewer than 150 reports across all channels of the full run of Fx94) data recorded via FOG in non-main processes might cause a crash. As a result of the crash, that data would not make its way to the data pipeline. This affects all platforms, but its rarity means it's mostly seen on Windows.
A fix is now verified in Nightly. I've requested uplift to Fx96 for it, and I expect that to come through.
However, some of your data from non-main processes in Firefox Desktop might not make it to the pipeline and into your analyses from clients running Fx94 and Fx95 (and early betas of Fx96). This should not materially affect any analyses I know of, and should not (upon being fixed) result in any noticeable changes starting in Fx96 in the new year, but if you have any questions or would like to have me look at your analyses specifically, please find me on Matrix, email, bug 1745660, or Slack and I'll be glad to help.