Take a soccer field with two teams and a ball on it. I would prefer to describe that as 23 particles of a few different kinds (ball, field player, goal keeper) and different properties (team membership, mass, fitness, whatever).
You could certainly invent a ball field and a field for both teams that are zero everywhere except where the ball or a player is located but you certainly wouldn't suggest that there are ball and soccer player fields in the universe even if they perfectly describe what happens on the playing field. On the other hand we could certainly agree that there are really players and a ball located somewhere on the field with specific masses, velocities and so on.
Maybe you can look at it from a point of redundancy. If you describe ball and players as particles with locations and momenta you get a description that naturally matches what is going on. If you use a field description you have to place a lot of constraints on what is allowed and what not, players disappearing, a second ball appearing and so on. The field description has just a lot more degrees of freedom than the particle description and you have to impose a lot of constraints to suppress field configuration that are not physically possible. So it also a kind of Occam's razor argument to pick a simple and minimal description.
I don't see how a "natural" description is any different from the word "real" here- fields describe things particles can't, especially when trying to use your "natural" intuition.
And by the way I am in no way suggesting any classical particle model, even without fields you still have the entire quantum mechanical machinery and there are certainly non-classical things going on. The question is whether you really need fields to describe some aspects of nature or if good old quantum mechanics is good enough and quantum field theory just makes things mathematical more accessible.