B'fanei Nichtav in Bavel 2

In this previous post we discussed Rashi in Gittin that interprets a gemara in Gittin to mean that ever since Rav came to Bavel, Bavel has a status of Eretz Yisroel as far as Get is concerned and there is no need to say B'fanei Nichtav in Bavel. The question on Rashi that we left off with was from the next line in the gemara. The gemara asks, what about our mishna where it says that even in Bavel there is B'fanei Nichtav?
According to Rashi, what kind of question is this? The mishna was before the times of Rav! The whole gemara seems to not make sense according to Rashi.
The gemara in Beitzah sets down a rule that a davar sheb'minyan need minyan acheir l'hatiro. Meaning, that if a Beis Din makes a takanah, the takanah can only be overturned by another Beis Din. There is a machlokes between the Rambam and the Raavad if, in a case where the taam hagezairah no longer applies do we need a beis din that is greater in wisdom and number (Mamrim 2:2). However, all agree that even if the taam of the gezairah is batel, some type of bais din still needs to overturn the gezairah.
Based on this we can suggest for Rashi that the gemara's question is based on davar shebiminyan. The gemara is asking that if Bavel was part of the gezairah in the times of the mishna, so how is that overturned in Rav's time. The gemara answers "l'vad mibavel", the gezairah was never made for Bavel in the first place.
Now, this pshat that I am suggesting is difficult based on the gemara in Gittin on 5a. The gemara there says that "l'achar shelomdu", after people learned the halacha of lishma, there was no need for b'fanei nichtav. This gemara is clearly saying that once the taam of the gezairah was batel, so was the takana. Apparently, the takana to say b'fanei nichtav was not a real gezairah in the technical sense, and the gemara in Beitzah does not apply to it.
So, now we are stuck in Rashi. If we use the idea of davar shebiminyan in Beitzah, the gemara on 5a doesn't make sense. And, if we don't use it, we are back to the question on our gemara.
As a side point, why wouldn't the takana of b'fanei nichtav be included in the gemara in Beitzah of davar sheb'minyan? We can suggest two possibilities. One simple possibility is that the takana in the first place included the "escape clause". For example, the takana may have been, "in places where they don't know about lishma, we have a takana of b'fanei nichtav". Another possibility is that the takana of b'fanei nichtav wasn't on the gavra like normal takanos. Rather, the takana was made on the get. The kashrus of the get is suspect based on the gezairah. Therefore, again, once the circumstances change the get's kashrus is no longer suspect and m'meilah the takana goes away.
The gemara by l'achar shelomdu ends off that we still keep the gezairah because "shema yachzor davar l'kilkulo", maybe people will again become ignorant. Rashi there implies that therefore the original takana stands (see sefer ayeles hashachar). The Sukkas Dovid explains pshat that really the gemara means that b'fanei nichtav was a davar sheb'minyan, and just because the taam is batel, still, the gezairah remains. So when the gemara answers that maybe people will become ignorant again, it's just a fancy way of saying that we aren't mevatel gezairos because the taam goes away... because who knows what the future will bring?
According to this, the sugya on 5a is totally consistent with the gemara in Beitzah of davar sh'b'minyan.
However, we aren't out of the woods just yet. This is because the gemara clearly says that kodem sh'lomdu the gezairah was stronger than l'achar sh'lomdu, in that kodem sh'lomdu the gezairah was even on milsah d'lo shchicha, even on uncommon cases. So, how can we say that it's all the same gezairah if the parameters have changed? For this reason many acharonim learn that they are two totally separate gezairos and not one. So what can we say for Rashi?
I would suggest that Rashi learns that kodem sh'lomdu there were 2 aspects to the gezairah. One on the kashrus of the get (cheftza) and one on the shliach himself to say b'fanei nichtav (i.e. on the gavra). The gezaira on the gavra cannot be batel, but the suspisions on the get can change with the circustances. Therefore, l'achar sh'lomdu on the gezairah on the gavra remained, which was more lenient.
Returning to our gemara, in Rashi we will suggest that this gezairah on the gavra is only in certain areas. The gemara thought that Bavel was one of those areas and included in the gezaira, so how could it be batel in Rav's time. When the gemara says l'bar m'bavel it means that Bavel never had any gezaira on it. See Tzafnas Paneach 7:10 who elaborates on this last point.