Yes, we have established previously that notes from consecutive reams in the same sheet position will be 9,000 numbers apart. Mixing of two reams in a bundle will produce two ranges of notes separated by 9,000.
It occurred to me that others have reported finding runs with missing notes such that it has proven quite difficult to obtain a run of 100 consecutively-numbered notes or even smaller consecutive runs within bundles. Perhaps random sheets are being pulled and inspected as part of a quality-control protocol, and when those sheets are deemed to be satisfactory, meaning they don't have problems that require culling, they are assembled and put together into a highly mixed up run of notes like the ones I found. As you can see in my grouping of notes listed above, the serial numbers span a range of about 250,000, or roughly 28 reams. Because reams are so small (200 sheets) and because there is no conscious effort to keep sheets in order before cutting, maybe it makes more sense to do inspections on randomly-pulled sheets. Maybe it is this kind of protocol that is responsible for the mixing of serial number ranges within bundles. For example, from every ream of 200 sheets, say 5 sheets get pulled, leaving 195. For every ream, there is a deficit of sheets needed to make 2 sets of bundles, so they borrow sheets from the next ream, and this "deficit" grows and, further down the line, more and more mixing of reams occurs.
If this is actually what is happening, it becomes that much more uncertain how replacements are deemed to be necessary in the entire note-issuing process. The pulled sheets are not replaced by anything... The printers are happy to leave gaps in the numbering. Perhaps one day, a brick hunter will get one brick of notes with a few notes missing, and he will get a second brick containing the missing notes from the first brick. That would be interesting to explain from the perspective of "insert notes".
