Show us the regular code, not the assembly output.
Why it works now is beyond me. The programmer won't override anything in PBP sends it or anything that PM sends it. The software that drives the programmer might, but definetely not the programmer itself.
In any case, I don't see a 7 second delay in the startup as a config word problem or something like that.
Maybe something in the circuitry just took awhile to charge up or power up and combined with no resistor on RB4 was the problem overall....who knows...
JDG