I'll add a couple notes to the nice answer by @SuperCiocia.
Regarding attractive vs. repulsive interactions. Your original intuition that you would want attractive interactions for a BEC is understandable. You want the atoms to be very cold and densely formed so that they bose condense. Surely attractive interactions would bring the atoms closer? This chain of reasoning is incorrect.
As SuperCiocia points out, the BEC transition is a non-interacting effect. If you had a non-interacting gas of bosons which was cold and dense enough it would form a BEC. Of course, again as SuperCiocia points out, in practice interactions are required for thermalization but that is a detail from the perspective of the phase transition.
So that said, we should understand that BEC doesn't happen because "the atoms are all attracted to eachother in a clump"*. Once we rid ourselves of this misconception we can then ask how does the introduction of interactions into the problem change things? The answer is short.
Strong attractive interactions cause the atoms to violently fly towards eachother resulting in inelastic collisions in which atoms can gain so much energy that they are ejected from the trap which is holding the atoms. The dynamics in light of attractive interactions is that you will just see rapid atom loss and no condensation. This is the collapse of the condensate.
What about repulsive interactions? A BEC can survive in spite of repulsive interactions. The atoms will be a little further apart than they would be without interactions but much of the major physics is unchanged.
So you should think like this: 1) theoretically our starting point is always a non-interacting BEC. Then we add in interactions. 2) If the interactions are attractive we get collapse. 3) If the interactions are repulsive things are changed/renormalized slightly but much of the essential physics is unchanged.
Rb-87 was more attractive for initial BEC than Rb-85 because Rb-87 supports an accessible cycling transition which could be used with early magneto optical trap and optical molasses technologies for laser cooling. Rb-85 does not support such a transition so more sophisticated laser cooling stages would have been necessary for the initial stages of cooling towards BEC. That is to say, Rb-87 probably wasn't specially chosen for its natural abundance
Rb-85 has attractive interactions and Rb-87 has repulsive interactions. For the reasons described above, this makes Rb-87 more favorable for realizing a BEC than Rb-85. Why does Rb-85 have attractive interactions and Rb-87 has repulsive? As far as I can tell the sign of the scattering length depends on the exact position of the continuum relative to the last atom-atom bound state, and whether the scattering length is positive or negative is basically 50/50 random chance for any given isotope. That is, it just happens to be the case that Rb-87 is more favorable with its repulsive interactions. The choice had nothing to do with natural abundance.
Why Rb among all other atoms? I can't speak much to this since I've mostly worked with Rb so far but I can point out that BECs of Na (for which the nobel prize was shared) and Li were formed shortly after the Rb BEC so I don't think we should infer something especially unique about Rb compared to other elements that it was the first to be condensed.
All of that said, I think your main question was really a confusion about interactions and BEC collapse which I think has been answered by now. It turns out that Rb has the right sign for interactions so that was helpful for it historic condensation.
*The question of why BEC does happen I will leave for you to research on your own or ask another question. The short story is that it is a thermodynamic transition that depends essentially on the fact that you have bosonic statistics and the density of states for the system.