Your perspective is backwards. Let's have a look. Why do people splash? A splash is advantageous when A: a class does not hold its own and B: the class is front loaded.
If B is true and A is not true, it will still be fully invested in.
If A is true then arguably B does not matter, because it is already a fail on its own. EK is a fail on its own. Therefore it is fail.
Intentionally designing something that does not hold its own is like making a leaky bucket with a hole on the bottom, to carry water somewhere. Perhaps if you dinged up your bucket by smashing a rock through the bottom, then you might be inclined to patch it back up. But buckets getting damaged is the wrong analogy here. If you were to
design a bucket for carrying water, why the hell would you design one with a hole in the bottom (EK), so that it would HAVE to be patched/combined with something else (another class) before it is serviceable?
Paths should hold their own AND provide synergy with other paths. There should be a point to being 100% EK, as well as a point to being EK+whatever. What you basically said is that the point of there being a the [EK] bucket is the handle. That we should just make off with the handle; screw the bucket.
The fact that it is splashable effectively, has merit. The fact that it is
only effective when lightly splashed, is a serious flaw.