The theory perspective
I know that for a $0Ω$ wire of uniform cross sectional area, the potential difference across its ends is zero
Yes but only trivially so. Why should there be any potential difference across you resistor at all? Is it in a circuit? Is there a potential across it? Is there a current flowing through it? You haven't mentioned any of this. A $0$ ohm resistor lying disconnected from any sources can have whatever potential we want at its ends depending on ambient electrical environment.
as no electric field is required to move charge with constant velocity (necessary to keep current constant throughout the circuit) as no resistive force is offered by the wire.
Of course there is no potential difference needed - you set the resistance to $0$! Your resistor isn't a resistor at all - its just a plain old wire.
But what if the cross sectional area of that $0Ω$ wire increases on moving? Do we still say potential difference across its ends need to be zero?
Yes, when in a circuit, definitely. Its dimensions don't matter. Its material doesn't matter. Its temperature doesn't matter. Once something has $0$ resistance, it cannot develop a potential difference across it with only a constant current passing through it.
Yes from Ohm's law it needs to be zero,
No. Ohm's law isn't valid here as $R=0$. The current going through a $0$ ohm resistor is independent of the potential across it, which as you yourself said, is always $0$. The current is determined by other elements of the circuit.
Real world perspective
but zero potential difference means that there is no electric field inside that wire, therefore no acceleration and constant velocity of charges.
Even though real world wires have some resistance and therefore some electric field inside them, even if they did not, charges won't be accelerated but neither would they be moving with constant velocity. They would thermalize quite quickly and obtain random velocities. If this was not the case, there would exist a spontaneous current (see Drude model)
But if charge continues to move with constant velocity and area increases on moving ahead then current through wire will increase. $I=neAV_d$
In a non-zero resistance with varying cross-section at equilibrium, the current through it is the same everywhere, its uniformity is unaffected by length or cross-section. This is because if the current entering and leaving a point in the resistor was different, there would be charge buildup/depletion at that point which is not something we model in a resistor.
Another reason the current won't increase with cross-ection is that even though there are more charge carriers, the electric field is smaller.
For your case of $0$ ohm resistor the equation isn't valid anyways since $V_d=0$ so current is zero everywhere.
In addition, the currently accepted answer states that
Actually the thing is ,ohm's law is not a fundamental law .Which means only certain conductor follow it...
Yes. Ohm's law isn't a fundamental law. But neither are most linear relationships like Hooke's law, Curie's law or other formulae that try to model phenomenon linearly. This phenomenological approach is both necessary and useful in describing nature. They are often simple and accurate enough for most daily applications.
However, simply because these laws aren't the most bare metal fundamental descriptions, doesn't mean that they are wrong in their regime of applicability. Such laws provide an approximate description, and when higher accuracy in predictions is needed, more refined, though often more complicated, models are used.
Ohm's law is very well followed by almost all conductors in the real world under ordinary conditions and by all in theory. Why its not being used in your question is because when Ohm gave his law, he was talking about the $IV$ relationship of things that do have resistance - not materials without it.
and OHM's law gives sort of average value of current
It does in the sense that the total current through the wire has to be a statistical average of the Avogadro scale charge carriers in a conductor. This doesn't make the $IV$ relation approximate or impart to it any error. Any other model would do the same. This isn't the reason behind your contradiction.
For a conductor of non uniform cross section ,electric field must be present to keep I along the length constant .
Cross-section has nothing to do with whether field must be present or not. An electric field is always needed to drive a current, constant or not. Moreover, presence of electric field isn't why the current stays constant along the length. As remarked earlier, it has to during-steady state to prevent charge-accumulation.
$^1$ there's another degenerate case: when its $\infty$.