Based on input so far, and conversations with Bill Troop, I wonder if these observations are due solely to film changes and do not lie with the fixer. In any event, the problem can be solved (according to EK) by increasing fix and wash times. So far, the only observations of what appear to be problems have been with Kodak film, right? The one observation about the Ilford film is a misunderstanding of the specifications which is natural given the test used.
The reason for confining it to Kodak film is critical (or may be here). Let me share an anecdote with you.
When designing TF-5, I was using Ilford MGIV paper and Kodak Polycontrast IV paper for paper fix and wash tests. That shows you how long ago this work started. Well, tests over 5 years ago showed that the Kodak paper took almost 2x longer to fix properly when compared with the Ilford paper. I had the same results with TF-4 and talked to both Bill and Bud about it. We had no explanation but then Kodak stopped making B&W paper. Well, at this same time, I got normal fix results from Polycontrast III and all other papers. Only the PCIV was slower to fix and we never had an explanation. I think that if Kodak had continued to make it, the data sheets for both TF-4 and TF-5 would read differently nowdays.
In any event, maybe something is odd going on here due to a change over the last few years, and maybe the 6' fix time suggested for some films in TF-4 and TF-5 should be applied more broadly than originally due to redesign of the films. A doubling of iodide level in a film can slow down fix rate by up to 2X. Dye layering can do the same as well as adding pink stain to the mix.
PE