I don't see why the debugging had to take place on 32bit binary. From experience, the debugger disassembly for 64bit is much more "readable" - as in much of the symbols called are spelled out a lot more, and accessing function arguments are much much more friendly - "po $arg1" instead of register math.
Hopper is capable of parsing 64bit disassembly and create pseudo code from that as well, if you need it.
Author of the post here: this was my first serious foray into disassembly land and I remember one post pointing out that Hopper had a harder time with disassembling 64-bit slices. However, it appears that was wrong! About to add an update to the post.
Hopper is capable of parsing 64bit disassembly and create pseudo code from that as well, if you need it.