The i386 failures, at least, should be fixed in 7.0.2 and later. Issue #220 upstream, affecting all platforms where time()+0x7fffffff overflows time_t. I haven't tried your 7.0.2+dfsg1, only my own packaging of 7.0.3.
Not sure about the hppa/powerpcspe and m68k failures, and I'm not in a position to test those architectures. I've found that debugging is a lot easier if one can see the detailed logs of the failing tests, as well as strace/ltrace output for the relevant command invocations.
