¬¬My 2 cents. Every collector is entitled to define "fancy" as he or she chooses.
I am in a unique situation. I have never collected fancy/geometric cancels, but
inherited several thousand off-paper stamps and 11-12,000 cut squares that were collected in a single small Indiana city, mostly between 1891 and 1893. I've been studying and writing about this hoard from time to time for nearly 20 years. The accumulation includes perhaps 2-3,000 machine cancels and probably nearly as man ellipses and ovals. Among the remaining thousands, there are very few really nice fancy cancels; there are many more of the ordinary manufactured and hand-carved fancy/geometric cancels. Some are really nice, clear socked on the nose examples, but many more are nondescript blobs. Most fall in between these two extremes.
Trying to match obliterator designs and determine the post office of origin in my case is a major challenge since I'm not working with entires. I haven't tried to count how many stamps and cut squares have identifiable post office information, but it must be a few thousand. That is largely because many cut squares were trimmed large enough to include portions of the CDSs. A few are large enough to include the entire CDS! I've been playing around and trying to compare multiple cancels from various post offices. For a single device in a small post office, the obliterator inking variations are readily apparent. Even over such a relatively short period of time (about three years in this case), the wear and damage to cancelling devices can be apparent, but that doesn't mean those devices are immediately discarded. For all we know, tartamimmi's example may fall into this category.
I also think some of the manufactured bullseye and target cancels, for example, have been modified. It appears to me that one of the rings sometimes gets damaged, and a creative PM or clerk simply removed the remainder of that damaged ring. Or, maybe they just wanted their killer to be unique and modified it. I suspect other modifications/repairs took place with the cork obliterators as they wore or were damaged. Variations appear, but it is difficult, sometimes impossible, to distinguish between wear, damage and uneven inking.
Aside from the widespread careless application of fancy/geometric cancels on the mail, my greatest frustration has been not being able to identify post offices of origin for some of the unique geometric cancels. Available resources include: the Cole book (now presented at Don's StampSmarter site); the Salkind book; older US Cancellation Club newsletters; and a dated online presentation from the Ohio Postal History Society. However, quite a few voids remain. The situation continues to improve and Don's StampSmarter project is a good example of how that is happening!