At least from a Rowbotham perspective, yes, the shadow object is a must. He also swizzles in the self-luminous moon bit as well. I haven’t read ENAG Chapter XI In a while. But enlightening:
"From the facts and phenomena already advanced, we cannot draw any other conclusion than that the moon is obscured by some kind of semi-transparent body passing before it; and through which the luminous surface is visible: the luminosity changed in colour by the density of the intervening object. This conclusion is forced upon, us by the evidence; but it involves the admission that the moon shines with light of its own--that it is not a reflector of the sun's light, but absolutely self-luminous. Although this admission is logically compulsory, it will be useful and strictly Zetetic to collect all the evidence possible which bears upon it.”
The last line is a particular favorite, kind of a, “Logic demands this conclusion,’compulsory’, in fact, but here on in, I’ll pepper you with cherry picked phrases from philosophers, astronomers, greeks and scriptural references to cement this undeniable logical conclusion.” type of argument. Clever, that Rowbotham.
Bottom line, the shadow object argument as a requirement to explain a flat earth lunar eclipse is just that; a requirement. One necessary to get FET out of the earth “getting in the way” jam. It is literally not based on anything observable, testable, measurable…nothing. It doesn’t even pass the Zetetic sniff test. It is a ‘compulsory’ necessity in FET, manufactured to solve a problem in FET that doesn’t exist in RET.