I'm going to disagree with Jim on almost all his points.
Point 1: Yes, it's true that the added height of the code 332 rail allows a bit more debris to fall between the rails and not cause derailments. I've operated on code 332 outdoors since 1980 and code 250 since 1996. I have found debris-related derailments to be no more frequent on the smaller rail. It's almost always something on top of the rails, not between them that causes problems.
Point 2: Code 250 is just as available as code 332,
unless you want sectional track. To my knowledge there is no code 250 sectional track available, though I thought I heard a rumor that one of the major players was about to introduce it. I forget where I heard that, so don't quote me. For flex track, you've got just as many (if not more) options for code 250 in terms of rail material and style of tie than you do for code 332. Sectional track, yeah--you're going to have to go code 332.
Switches in code 250 are available in as many different sizes as they are in code 332, though not quite as small a radius. The smallest non-custom-built code 250 switch manufactured is a #3 switch, which is nominally a 30" radius.
Point 3: Not the case. Code 250 brass flex track can be found for around $3/foot where code 332 will run you around $5. Brass sectional track in code 332 is now upwards of $6/foot for the shorter sections. Code 250 switches cost more on the tighter end of the spectrum, but when you compare wide-radius switches (#6 or greater), they're within a few dollars of each other. The bigger cost comparison will have to be done relative to a particular tie style and rail material. (i.e., not everyone makes stainless steel track, so you're going to be limited to pricing your choices for that particular material.
Here's my take...
Advantages of code 250 rail:
1) More realistic (scale) appearance.
In 1:20.3, code 250 rail comes in at around 5", which is equivalent to 80# rail, typical of heavily-traveled narrow gauge railroads such as the D&RGW and the EBT. In 1:32, it scales out to 8", which is the same as the 155# rail used by very heavily traveled standard gauge mainlines. For scales in between, the rail still falls well within the range of scale appearance.
2) The rail is easily bent without the aid of a railbender. Of course, with the advent of the 2-rail railbenders, I'd never build a railroad without one again, but you can definitely "belly bend" code 250 rail with little difficulty, at least to wider radii. Below 5' radius, you'll want a railbender regardless.
3) Code 250 rail is just as strong as code 332 rail, provided the support beneath the track is solid. Most of us have seen the old LGB ad with the elephant stepping on their track to show how durable it is. The fact of the matter is that the actual pressure being put on that track by the elephant's foot is less than that put on the track by a person if they were to step on it. In truth, N-scale track would hold up equally well if an elephant stepped on it, provided it was sitting on an asphalt slab that didn't give beneath it. That's why rocks and twigs don't get crushed when an elephant walks on them. They may break at the ends, but that's because the ground is soft and there's a shearing effect. If you want to showcase durability, show a photo of a woman in a stiletto heel walking on your track.
Disadvantages of code 250 rail:
1) (as Jim mentioned) The smaller rail means larger rocks, acorns, twigs, etc., that fall between the rail have a greater chance to cause derailments. Speaking of my experience only, this has never proven to be a significant issue.
2) Not available in sectional track. Personally, I'd put this under the "advantages" column because I think sectional track limits one's creativity when planning a railroad, but if you're going to use sectional track, you're pretty much limited to code 332.
3) The smaller cross section of code 250 is not as strong as 332, provided the rail material is the same. Stronger metals in code 250 may be as strong (or stronger) than weaker metals in code 332. Again, with a firm foundation supporting the track, this is a moot point anyway.
Common Code 250 myths debunked:
Trains need smaller flanges to run on code 250 track. Not the case at all. Code 250 means the rail is .250" tall (1/4"). The deepest flanges on commercially made rolling stock fall right around the 1/8" mark, and standards in use by various model railroad organizations (G1MRA, etc.) specify even shorter flanges. The spike details on code 250 tie strips do not come close to exceeding 1/8", so the flanges will operate just fine.
Code 250 is incompatible with code 332. Just like the prototype, there are "adapter" rail joiners which allow the different sized rails to be joined so that the transition is smooth and seamless. If there's a particular piece of track that is only available in code 332, you can use it and buy a set of such railjoiners for each end of the track section.
Code 250 doesn't hold up outdoors. This goes back to the strength issue, which is simply not applicable. If the foundation of the track is weak, then yes, the code 332 has some advantage because the stronger rail makes up for more deficiencies. Personally, I don't want a weak foundation under my track; I want to know my locomotives aren't going to roll over on their side. If something catastrophic were to happen to the track (a tree limb, large rock, or large wildlife with small feet such as deer or elk) then the size of the rail isn't going to make a bit of difference.
The bottom line is that in terms of operation, there is no area of garden railroading where I would say that code 332 rail would be a "must" over code 250. They are--in terms of trains rolling over them--every bit equal. The choice between one vs. the other hinges foremost on whether you want to use sectional track, then cost for your particular choice of rail and tie materials from one manufacturer vs. another. For instance, if you want stainless steel rail, your choice of manufacturers is limited to a small subset, so you have to look at the cost of just their individual product lines to see which is cheaper. Don't discount aesthetics, too. If you're modeling 1:20.3, then something like Aristo's US standard gauge ties are not going to look near as good (narrow gaug-y) as those specifically made to represent narrow gauge ties. One may be cheaper, but if you're looking for a specific aesthetic, then the extra cost is well worth it.
Later,
K