The preferred acronym for what your referred to as "single-ended parallel" is PSET, "parallel single ended." SEP usually is intended to mean "single-ended pentode." Of course this convention collides when SET, usually intended as "single-ended triode" is instead used to refer to "single-ended tetrode."
There's nothing wrong with a well-designed and executed PSET circuit. You need precise matching of components in the circuit to maintain the incisiveness and purity of a simple SET implementation of the same tube, but these are matters of execution that *may be* expensive.
A poorly-executed PSET circuit will introduce some blurring of the sound spatiality and some subtle smear in transient definition, compared to the same design in a simpler SET implementation using one tube of the same choice. A poorly-executed PSET circuit using 300Bs can also fail to manage the rising deep bass distortion of that tube. On the other hand, my Audion Golden Dream PSET 300B monoblocks clearly outperform the same company's excellent SET 300B monoblocks that are one step down in their line, within the power limits of the lower-output amps, including delivering the best deep bass I've heard from a single ended amp using the 300B tube. If you don't like PSET compared to SET, you just haven't heard the parallel instance executed to a sufficiently high standard.
Phil