I don't doubt that Germany does a very good job at recovering photo silver, but I'll say that it's very difficult to get at that last couple percent. And in fact, it's very difficult to even be exactly sure where you're at. For some reason, silver content of materials is closely held information by the manufacturers. I've seen published information on certain limited films (it's confidential), but it's always referred to as "recoverable silver content."
Here's an example of why the last bit is very difficult. If your process uses normal amounts of wash water, the wet film exiting the fixer tank brings a small amount of silver out with it, then lets that silver spread into a large amount of wash water. One way to chase that silver is to put your wash thorough ion-exchange columns with specialized ion-exchange resins. There will be some slight leakage of silver through the columns. Then, when you strip the columns and attempt to recover that effluent, some silver will slip through. Likely, some silver will end up trapped in the resin beads. So you have trapped it, but you won't have any idea how much is there until the resin end-of-life, when you ship it to your refiner.
If you want to make things a bit easier, you could use the so-called washless sytsems, with very low flow rate systems. They keep the silver in a smaller volume, so more concentrated, where it's easier to get at. The downside? Probably the film is not as thoroughly washed, so more silver remains in the film (and the customer keeps it). A way around this is to have a processor with a large number of counter-flow tanks, etc, but such machines would be more expensive.
Anyway, it's very difficult to get at the last bits of photo silver, and probably not economical. So there is a tendency to only go as far as necessary to meet the effluent requirements.
Sorry if I'm boring you too much; it was always an interesting challenge to me.