The Order of Things

I'll spare you my more esoteric ruminations on systematics in favor of the following concise-ish points:

1. It is amazing how accurately the Linnean system has been able to predict the actual genetic relationship between taxa.
2. Its failures in this regard are increasingly apparent, however, as in the decimation of the family formerly known as Scrophulariaceae by molecular phylogeny, noted above.
3. This is a failure not of the Linnean system per se, but of the attempt to conflate one system of representation (flower morphology) with another (phylogeny).
4. But both systems are "true." The latter maps its categories onto the empirically verifiable genetic relationship between plants, so is perhaps more useful, or more revealing of the underlying processes of evolution. But the morphological similarity between, say Antirrhinum (moved to Plantaginaceae) and Verbascum (still in Scrophulariaceae) is still absolutely "true" (as long as it is stripped of its phylogenetic pretensions).
5. Horticulturally, of course, this is neither here nor there. The categories relevant to gardeners more closely resemble Borges's (fictional) Chinese encyclopedia than either of the above: o. plants that remind me of something pleasant when grouped in a certain corner that I often neglect to water.
6. This explains the historical reluctance of gardeners to deal with taxonomy. It's not just laziness, it's also because leaf serration, or water requirements, or, especially, flower color, all more or less excluded from the Linnean system, are vastly more important to them (us) than, say, the relative glabrousness of the peduncles.

That's all. Go read Borges again.