Good post cjam.
I think the crux of the matter is: "Is the approach under control?".
If you're at idle thrust, 1.5 dots high and hanging on the speedbrake at 500'/1000'/whatever, then probably not and throwing it away would be best.
If the speed is not quite where you would like like it to be but corrections are being made and are having an effect then continuing is a possibility. Obviously, if you're outside your company specific limits at your 'gate', then carrying on is not a good idea. On the other hand, if it's just turbulence/wind gradient/thermal motion causing a momentary departure from normal parameters, I don't see why that should cause undue alarm, as long as you're doing something about it.
I'm glad that there has been a 'sea change' in culture from "Yee-Haw!" sideslipping glide-ins with top-ruddered final turns to the 1000'/500' ARTE stabilised approaches we all aim for today. At the same time, we shouldn't let it become a 'numbers game' and should try and remember some of the reasons why changing to stable approaches was such a good idea:
* Power up before landing = engines in the rapid acceleration range, so able to cope with unplanned G/A, etc.
* Excess speed over threshold + long flare = deep landing & possible overrun.
* Stable airspeed, pitch & power make it easier to detect the onset of windshear, i.e. one or more of these parameters is behaving unusually.
* Both guys are 'in the loop' with the aircraft operation, as opposed to the 'fighter pilot' style of fast (rushed) approaches.