Seemingly there would be room to distinguish between a definite ruling and doubt, since the law of majority causes us to treat the half that separated from the shop as definitely kosher, in contrast to the fixed half, regarding which “anything fixed is treated as half-and-half,” and therefore only from the standpoint of doubtful conduct would we judge it as forbidden.
Because of that, there is no side saying that the separated part should follow the fixed part, since the separated part has the conduct of certainty, whereas the fixed part is a doubt. There is only room to discuss whether the fixed part should follow the separated part, since the separated part is definite, or perhaps “anything fixed” applies here too, because so long as there is no clear decision it remains in its doubt.
What do the rabbis think of what I’m saying?

And by the way, if so then the size of the separated part makes no difference—it will always remain permitted.

I’m not sure that fixed status is not a law of certainty. For example, Rav Elchanan Wasserman writes in an article that all Torah-level doubts where we rule leniently (mourning, mamzer, firstborn, impurity in the public domain, orlah outside the Land of Israel) are laws of certainty and not doubt.

And on the other hand, the claim that majority is conduct of certainty is by no means universally agreed upon. Later authorities disagreed about this regarding majority in impurity in the public and private domains—whether it is in the category of doubt or not.

You mentioned the Golden City, where the Talmud says (“and it shall be found written,” Sabbath 138a) that we do not find a dispute from one extreme to the other—one permits ab initio and one makes one liable to bring a sin-offering—unless there is an intermediate opinion: exempt, but forbidden. Tosafot remain with the difficulty: what does the intermediate opinion help, since in the end there is still a dispute from one extreme to the other? Between the lines it sounds as though you have an explanation of that Talmudic passage (because usually one brings proofs from things one understands; otherwise who says that after the real explanation—which is needed anyway—the proof will still stand). If so, could you share it?

As for the substance, I didn’t fully understand your point in bringing it. I understood your line of thought like this:
I was wondering whether such a possibility of this kind of moderate pluralism exists—a single decisor who holds several contradictory reasonings all together, and not merely as doubt, but still one cannot just choose whatever one wants; rather, one counts them for the purpose of majority. If so, we would presumably have to find such phenomena explicitly elsewhere too, because in many disputes someone presumably thought that “there is room for all the reasonings.”
And you replied that these are not differing opinions but sides in reasoning that are present and operative, and following sides is something familiar, like the Rashba on double doubt. [Still, there the decision is in the laws of doubts, and the truth is one and known before Heaven.] You further replied that we find something similar in judicial discretion between litigants [though there too it is because we do not know the truth], and in “one who acted like this master acted” between sides [though there too it is seemingly by the law of doubt and the permission to rely on one’s primary rabbi].

And in the course of this you also came to answer that one should not ask “where do we find such a thing?” (apparently because the answer is: here, we do find it). To this you brought the Golden City, where the Talmud asks where we find such a pattern of dispute, from one extreme to the other—and that seemed to support my surprise: where do we find such a pattern of opinion, namely this moderate pluralism like the Chazon Ish? But then you added, “and it implies that one Tanna determines his opinion in light of the other Tanna” (and in parentheses you raised another possibility), and I didn’t understand that addition or what it was getting at.
So did you only bring the Golden City to show that we do find questions of the type “where do we find such a thing?” And then answer that here too, in the Chazon Ish, we do indeed find the principle of following sides in several different forms—majority of sides, and Rabbi Shimon Shkop in the case of shops, and judicial discretion, and “one who acted like this master acted”?

I brought it as an example that Tannaim formulate a position through relating to the positions of their colleagues. If I see that one says a sin-offering is required, then I, who disagree with him, will not permit it but will forbid it rabbinically. But as I said, that can easily be pushed aside.
As for the explanation of that Talmudic passage, there is a strange proposal based on quantum logic in an article by Daniel Weil in Higgayon, issue 1. Simply put, it seems that if there is an intermediate mediating opinion, then the gap can be large. It is not reasonable that there should be one opinion that permits and one that forbids at the Torah level, with no middle opinion that forbids rabbinically. But if there is an intermediate opinion, then even polar opinions can exist.
By the way, that passage suggests that an opinion that forbids rabbinically is apparently based on the idea that a rabbinic prohibition is half of a Torah prohibition—that is, there are reasons to forbid it at the Torah level, but they do not cross the threshold, and therefore only the sages forbid it. This is unlike views that treat a rabbinic prohibition as transparent relative to Torah law and wholly a novelty of the sages. Similar to the well-known distinction between two kinds of rabbinic Sabbath restrictions.

I looked into the Talmud there in Sabbath 138a and noticed something.
A. From the language of the Talmud it seems that what allows a far-reaching dispute is not the mere existence of an intermediate position; rather, an intermediate position allows us to say that each side argued and discussed with the intermediate position, while the distant extremes did not argue with each other at all. Rav Yosef brought a proof from the Golden City that an extreme dispute is possible, and Abaye answered him: “Did you think Rabbi Eliezer was responding to Rabbi Meir, who said she is liable to bring a sin-offering? Rabbi Eliezer was responding to the Rabbis, who say exempt but forbidden, and he said to them that it is permitted ab initio.” It sounds as though Abaye was not telling him anything new about the fact that there was an intermediate opinion and the significance of that fact—that was obviously known to Rav Yosef, and it is not what Abaye’s wording suggests. Rather, he was claiming that the two distant sides were not arguing with each other at all; each side encountered and debated the intermediate opinion.

B. In the Golden City there is a far-reaching dispute: Rabbi Meir requires a sin-offering, according to the Sages one is exempt but it is forbidden, and according to Rabbi Eliezer it is permitted. There is no necessity that “the Sages” are one distinct group. It could be that in Rabbi Eliezer’s generation there was one group of sages that disagreed with Rabbi Eliezer—Rabbi Eliezer says permitted and they say exempt but forbidden—and in Rabbi Meir’s generation there was another group of sages that disagreed with Rabbi Meir—Rabbi Meir says liable to bring a sin-offering and they say exempt but forbidden. And all those sages were grouped together under the single label “the Sages” in the Mishnah.

If indeed the whole problem is that two Tannaim should encounter one another and still remain in a far-reaching dispute, then that changes the picture somewhat. And it strengthens the possibility you suggested (and dismissed as easy to reject) that Tannaim relate to one another’s views—that is, to one another’s reasoning, and therefore it is not likely that they would hear each other’s reasoning and still remain in a polar disagreement. Presumably that would happen only when the dispute is polar at the root, and not merely that polar outcomes are drawn from the root. [There would still remain proof that the rabbinic category is an intermediate domain between permitted and forbidden, and that there is a continuum here. But perhaps there is also another explanation connected to tradition and custom, especially in something relatively common like carrying on the Sabbath. Usually the sides do not want to confront tradition and custom directly, even if they have the power to do so. The explanation would be that two close opinions have common ground, so perhaps that common ground is what was transmitted in tradition, and within that framework they disagreed; and the local custom could not decide because it was probably borderline. But if two distant opinions disagree and encounter one another, they will discover that there is no shared tradition and they will have to do some kind of epidemiological investigation to discover where the communication breakdown arose; and perhaps one could also decide from custom. But if they did not encounter each other—that is, they were in distant places and different times—then one could say that the custom in each place remained as it was, and as for the corruption in the tradition, they would not know where to begin clarifying it. But that needs more development, and perhaps it does not fit the sugya.]