Ilford is just talking about pushing for speed. This is no different than how Kodak’s P3200 or Ilford’s Delta 3200 is approached. Neither has an actual ISO because it can’t with the extended processing and having an ISO just for normal would probably confuse people. Just remember EI is the setting on the meter and not film speed.
From Kodak’s P3200 Data Sheet, “The nominal speed is EI 1000 when the film is processed in KODAK PROFESSIONAL T-MAX Developer or KODAK PROFESSIONAL T-MAX RS Developer and Replenisher, or EI 800 when it is processed in other KODAK black-and-white developers. It was determined in a manner published in ISO standards. For ease in calculating exposure and for consistency with the commonly used scale of film-speed numbers, the nominal speed has been rounded to EI 800.”
View attachment 405496
From Ilford’s Delta 3200 data sheet,
View attachment 405497
The ISO prefix can only be used if the conditions of the standard are adhered to. This means testing a large number of rolls over a period of time using multiple batches. It's a laborious and expensive process. The manufacturers are going to limit this level of testing to one or two developers per film. All the other film / developer combinations can be tested using the parameters of the standard minus the large sampling and consequently can't use the ISO prefix. They obviously would also want to do a family of curves for as many combinations as possible. They want to know what their film can do and for the consumer to get the best possible results. The standard only has one set of contrast parameters, so ISO can't be applied to anything processed outside those parameters. But the manufacturer can adhere to how the standard on how the film is exposed, hold times, and such and the film can be assigned an effective film speed. Even with P3200 and Delta3200 data sheets are for each manufacturers’ film, it’s not certain they did the complete ISO protocol.
The data sheets are a balancing act as they need to convey the technical information of the film, but in a way the majority of the general public can understand. Because of this, those with more technical expertise can sometimes get tripped up by the lack of precise language. Ilford uses ISO and it is clearly displayed in the data sheet and on the film box, but the data sheet talk about processing. Here's where the sheets are geared more to the general user. Rather than speed being a product of the characteristics of the film in a given developer at a given gradient, most people use speed as pliable and base the development on the EI rating. Which is technically wrong. When the manufacturer offers a development time based on the EI rating higher than the the film's ISO rating, they know the extended development won't increase the effective film speed to the new EI rating, but will restore density range which is most likely lost because of underexposure. Kodak’s P3200 data sheet, “When you need a higher speed, you can expose this film at EI 3200 or 6400. At these speeds, there will be a slight increase in contrast and graininess with additional loss of shadow detail.” In other words pushing for speed.
As far as I can tell, the main difference between the two manufacturers is that Kodak appears to use the rule of 2/3 stop contrast for every stop underexposure, where Ilford appears to use field testing to confirm acceptable results (although there is no way they don't know the sensitometric values or haven't run family of curves on all the different films).
Kodak actually includes their CI values in their Xtol data sheet:
View attachment 405503
And from Kodak’s Tri-X data sheet:
View attachment 405500
How is it any different than Ilford’s? Kodak didn’t list the ISO processed in other developers and this is their own film and developers. What most likely happened with the OP’s FP4P experience is experimental error, but it’s not unheard that the data sheets contain errors. It’s hard to compare recommended development between Kodak and I Ilford for a given film because Kodak rarely acknowledge other films and developers existed, but the Xtol data sheet is different. Kodak has FP4P as 8 min. Ilford has it as 8.5 min. Their respective average gradients for normal development are 0.58 and 0.62, so that lines up.