Both your objections are true, and I agree. If NS is explained in this way it is not a tautology.

But this line of reasoning is also anticipated by the author who's views I'm not testing.

His responce was that this line of argument leads to a definition of 'fitness' (though you did not use this word here, so maybe I should say 'selection criteria', that is essentially so complicated that it should be cassified as metaphysical.
Point is: there is always a selection criteria as you point out. Once you take away one selection criteria there will be another and another and another, but it is very hard, if not impossible to predict beforhand which criteria will have the upper hand once the most obvious ones are taken away.

He compares it to astrology where the stars always predict the future. If not the stars, the season, if not the season, the positions of the planets, if not the planets, the moons of the planets, or the cyclical period of the sun or whatever.
How to choose from all of these? Take a single example and find the cause.
How to choose on what basis nature selects today? Take a slingle example and you'll find the cause.

It is fundamentally untestable when applied to nature as a whole since the criteria for selections are explained to be "very complicated" and "virtually inpredictable" and hence meta (beond) phycical.