Mike Cross,
Yes, looking back at my post, my wording was wrong. however, you seem to be forgetting something. Firstly, as you say, the almost closed throttle plate acts as a second venturi, i.e. a further narrowing of the intake pipe. The pressure drop here is much greater than in the main venturi, as the remaining orifice across the throttle plate is very small at idle. Any ice forming here can very quickly block the inlet path.
BUT.. don't discount the cooling effect of fuel flow through the carb. Liquid fuel issuing from the fuel jets is merely atomised, i.e. in aerosol form. Evaporation of the fuel droplets may NOT be completed until after it has entered the manifold, so cooling from this effect (pulling in latent heat of evaporation) continues right through the carb, definitely including the area around the throttle plate, taking heat from it. As the plate has a small mass and can only draw "replacement" heat through its spindle, it cools rapidly.
This is why most cars with carburettors have water-heated inlet manifolds - to ensure the fuel is finally and fully converted to vapour before it enters the cylinders, as liquid fuel cannot burn. Remember how a fridge works? The cooling in that case is brought about mainly by evaporation of a liquid. Your "venturi only" theory would suggest that refrigerator manufacturers have been doing it wrong all these years - perhaps they only need a suction pump and air flowing through a venturi - wot no harmful CFCs?
I once had a motorbike that could actually ice up its carburettor without the engine even running. It had a leaky float chamber valve, fuel sometimes came seeping out and when left for a while in suitably humid conditions, ice would form on the outside of the carb - no pressure drop there! Interestingly, the ice looked like pink cake icing as it was mixed with the colour of petrol and 2 stroke oil. It only did this in summer, which ties in with advice given about the worst conditions for aircraft carb icing in UK.