pachos pachos m'daled amos (Eruvin 34a)

I just learned today's daf with Rashi but there were a couple of points on which I was confused.

1. At the beginning of the gemara Rashi says that even if you have to go through reshus harabim to get your eruv to your makom shevisa that's OK since you could do that bein hashemashos by carrying less than 4 amos at a time. Then at the bottom of daf 34a, he says in his second pshat when explaining the gemara which says that if you lean the shovach or migdal over it will be more than four amos away that the problem is that it's more than 4 amos in the reshus harabim. Doesn't that contradict what he said earlier? The Maharam answers that our imagination can only run so far. In other words, we can allow you to say that you could get it bein hashemashos if you carry less than four amos at a time and we can also allow you to say that you could get it if you lean the shovach or migdal over but we can't allow you to say both of them at the same time. That requires way too much pretending to allow the eruv. (I just checked the insights to the daf and they quote the same Maharam in different words and also have more to explain Rashi's pshat)

2. The first thing that bothered me about the case of the window and the rope is what changed between the hava amina and the maskana - didn't the gemara think that it's likely the guy didn't have the rope with him? Tosafos asks another question: Why all of a sudden here did the gemara introduce this idea? Why couldn't this have been possible in the case of the tree from daf 33? Tosafos answers that just isn't common so the gemara didn't even entertain the possibility. This would answer my first question - the gemara assumed that most people who are on top of cabinets next to walls have ropes with them (are you going to argue with that?) but the gemara answers that's not true. He also suggests another answer that since we're already saying hoil here we ask this question but in reality we could have asked it earlier.