Does the starter speed vary in a pattern, like a musical beat?
If so,
is it consistent with the revolution of the starter motor?
( If so, suspect starter commutator/brush problem . Did fuel or oil get IN the starter and glaze ( like burned-on dinner in the frying pan ) on the starter commutator/brush assembly? Fluid leaks seem to be attracted to starter motors, as they are usually mounted at a low point. )
Is it consistent with the revolution of the engine?
( If so, suspect a cylinder compression issue. Your boyfriend mechanic would probably recognize this "melody" immediately. ). Hope it's not a head gasket or cavitation thing.
If no pattern is noted ( human brains are excellent pattern detectors - it's music in a way ), I second Jesus Freak's advice. I've seen the exact same thing he described. The battery cables look and may even feel intact, but inside, it's just corroded copper where years of battery acid have wicked into the cable and ate the copper.
Personally, I like a small battery powered oscilloscope for checking out this sort of thing, as it shows quick events and freezes them. It's kinda like comparing a photo to a movie.
This is what I use...
You can electrically verify the connection by measuring directly to the battery terminals, + and - poles, before and during start, then look at what shows up at the starter motor.
It's a 2-person job...one to fiddle with the key, and the other to fiddle with the scope and probes, as it's really risky to fiddle with clips around starter motors...one slip and, well, you know...flying molten copper.
I find it a lot easier to measure, then wrench.