Upload & Sell: Off
I have no problem with the cost/value test as each photographer has their own budget and know what they want to prioritize. It's the "degradation" argument that doesn't hold up IMO.
The argument against protective filters is not based on any single issue, such as degradation of the image, but on the overall question of the trade-offs involved. A rational discussion of the issue doesn't focus entirely on image degradation (or one would never use any filters) or cost (since cost is often measured against value) or anecdotes (since some will almost invariably support whatever position the speaker favors), but on the overall question of whether, on balance, these things are providing the value that some imagine and at what cost.
A few high points:
1. While there are stories of a filter minimizing or preventing front element damage in the case of something striking the front of the lens, there are also stories of an impact that would have created little or no lens damage instead breaking the filter into pieces such that the shards damaged the front element. Choose your poison.
2. While the filter may prevent some kinds of lens damage, its insurance value is limited to the small percentage of ways that a lens can be damaged that involve an object striking the front element. They won't help, for example, with a dropped lens except in very rare cases.
3. While there are some cases in which a filter might minimize some sorts of front element damage, the vast majority of photographers will never have such an accident - it turns out to actually be quite unlikely, and the risks can be minimized even more by adopting smart ways of handling lenses.
4. The UV filtering capability of the filters most often sold for protective purposes is entirely worthless on modern DSLR cameras, which are not subject to the effects of UV light in the ways that film supposedly was.
5. The extent to which the full-time use a filter might affect the image quality varies - this is not a simple "yes it does! no it doesn't!" proposition. The use of a very high quality coated clear-glass protective filter will not produce any noticeable degradation of the photos in the large majority of cases. However, the use of inexpensive protective filters (of the sort often sold as part of "kits" or purchased by those who can't see spending $100 or more on a filter) is much more likely to have negative effects in normal shooting. There have been tests showing reduced contrast, additional reflections or flare, a softening of the image, and even optical distortions from cheap filters. Even with the best filters, internal reflections from the back of the filter can cause "ghosting" effects in night photographs and similar work in some cases.
6. The cost benefit story is not as simple as "it is worth $100 to save a $1200 lens!" For example, this makes the idea of using a protective filter on an inexpensive lens pretty silly. The value of the lens might be not that much more than the cost of a filter good enough to avoid the optical issues described above - so not using filter makes sense on a $100 or $200 lens from this perspective. Putting an inexpensive and less than optically great filter on an expensive first-class lens doesn't make much sense, and excellent filters for, say, 77mm filter diameter lenses, can cost in the range of $100 each. You can easily spend $400 or $500 (or more) to outfit a set of excellent lenses with individual filters. Unless you imagine that you'll damage all of them in ways that would have been prevented by filters, the cost benefit starts to look a bit less interesting.
7. In the process of preventing possible (though not always certain) front impact lens damage, the filter will be destroyed or at least rendered unusable. So while you can imagine that you saved the cost of the lens - though actually, in many cases, you saved the much lower cost of a front element replacement or of living with a small scratch - you lost the cost of the filter and you'll now have to replace it.
8. The very sorts of damage that a filter might protect from can also be warded off by using the lens cap you already have when you are not shooting - and this provides better protection than the filter - and/or a lens hood when you are, especially with long focal length lenses. (The hood provides considerably less protection on wide angle lenses.)
9. Related to all of these points and especially to #8, the filter will not protect from all foreseeable front impacts, much less protect from a wide range of other types of damage such as dropped lenses/camera, side impact, etc. So it seems to me that anyone who is so concerned about supposed lens protection from a filter would also invest in a serious insurance plan to cover all of these types of damage and other real risk such as theft. Do you?
10. There are undoubtedly a few cases in which using a protective filter makes sense. If you regularly shoot in very risky environments (and this doesn't include occasional shooting along the beach or in the desert nor probably the potential of finger prints from a child touching your lens... ;-) and you work with a camera body that is sealed and can therefore also be protected from such threats, adding a filter can seal some (but not all) L lenses and complete the relatively greater protection. (I'm often amused when people who shoot a relatively unsealed DSLR body insist that their lens must be sealed...)
In the end, everyone has to make his or her own decision about the value and costs of using protective filters and about employing alternatives ranging from careful handling, to using filter/hood, and buying real insurance. But there is nothing wrong with thinking about the real costs and value of the various approaches before you spend money on any of them.
View previous versions