Actually, filtration is a very common CCP and is measurable. Watanka's post (#4) explains it pretty accurately. If you really want to challenge your filters, aside from putting foreign objects in the line (not my favorite tool), you can also measure the sizes of any particles captured by the filters. In this case, I presume that might be more difficult for the types of particulate captured in the OPs process, but in other products with a coarser filter, this can be achieved.
We declared last filtration before bottling to be a CCP (physical) for bottling distilled spirits. Our filters run from 25-micron to 0.5 micron depending upon the product. We check the filters at least daily, more often if there is a change-over to another product type. QC Technicians inspect the filters for visible tears and holes. FDA declared a limit of 7mm for physical hazards per CPG Sec. 555.425 Foods, Adulteration Involving Hard or Sharp Foreign Objects. We set critical limits to no visible defects in the filters. Human eye can resolve to about 35-40 microns, so a visual inspection is more than adequate to that task. We document filter inspection observations. We conduct challenge studies at least annually as a part of our HACCP reassessment using a 6.5mm object to ensure filtration catches it in each bottling line. Check your UK regulations to see if there is a physical hazard size limit like FDA's 7mm limit for physical contaminants, or just set a critical limit of no visible defects.
Mr. Watanka clearly mentioned (marked in red-above) which says it is visual inspection and i dont think he has mentioned anywhere measuring of the CCP except annual challenge test (which we also do for 20 mesh sieve in my Organization). Big task is 5um filter!!!.
Currently we are measuring the 20 mesh sieve at random places every 4 hours and limit i have set for this CCP is 0.841 mm to 0.905 mm. We also have metal detector before sieving which i made it as OPRP.
CCP is fine if the filter mesh size is bigger (below 20 mesh) i.e. openings above 0.841 mm. But the filter mesh size is too small such as 5um, then it is very difficult to measure and also it is time consuming.