The one caveat I can identify is that the typical scientific process will include release of a paper into the scientific domain, where it should be scrutinized by the community at large. Anything that doesn't stack up should be shredded, and having to revoke a paper can be a crippling event to any scientist's career.
One of the problems with the free market model is that science must be proprietary, not shared, and kept hidden for as long as possible. Medicine has an uneasy balance, where tradition and law require sharing of information and peer review, but science in industry for such things as genetic manipulation of crops does not have any such mechanism for peer review or sharing of information.
On the topic of GM foods, I agree that 'crop husbandry' is a far cry from genetic manipulation of crops the likes of which we see today. But to the best of my knowledge, there's little danger to eating such crops. The genetic composition of a foodstuff doesn't pose a threat to us as biological organisms: there is no chance of an abberant, artificial DNA sequence 'infecting' our cells and causing cancer, for instance. Genetics simply don't work like that.
On the contrary...
First, it is not so much what the genes inserted into the crops will do to our genetics, it is how they express themselves in unknown ways in the crops or food animals in which they are inserted. Prions, as I'm sure you know, give rise to incurable and untreatable diseases in some cases, and they're nothing more than proteins. We know that huge sections of genetic material once thought of as 'dark' may well have functions we simply do not understand yet, and they may express themselves in a variety of ways.
Second, I just read yesterday a news report about a study which shows that parts of our own DNA are composed of ancient retrovirii which self-destructed in reverse-transcribing their RDA into their host (human) DNA. What does it do? No idea. Could be benign, but some endogenous retroviruses have been suspected in certain cancers. Some companies that go GM food engineering are experimenting with RNA silencing as well, not to mention intron manipulation - which gets us back to exons and yes, prions producing proteins once more. Introduce a retrovirus? Who knows?
More of a concern is what the gene sequences code for. If a crop was designed to secrete a weedkiller that it was immune to, then this could have massive reprecussions for us. We'd be exposed to weedkiller every time we ate the plant. On the other hand, fortifying grains with vitamins shouldn't pose any threat at all (provided that in high doses, the vitamins aren't carcinogenic or any such problems.)
There are secondary issues. An example is the weedkiller resistance. As this occurs, farmers are able to dump massively increased amounts of weedkiller (primarily Roundup) on their crops without damaging them. This increases runoff and pollutes watertables. It even makes it into our drinking water in many places. Hope you like weed killer, you're possibly drinking it.
Despite occasional public news-bytes that tell us how GM food is going to put things that are healthy for us (omega-3 fatty oils) into food that would normally not have any, most GM engineering is designed to satisfy market requirements. Food is engineered primarily to resist pesticide application, resist typical pests or diseases, stay fresh-looking longer, be more resistant to damage during transportation, and taste better. There is no capitalist impetus to make the food better (healthier) for us. That would only come from government regulation, and the government doesn't (in the US, anyway).
Genes aren't a threat: what they code for could possibly be. As long as we're sure that the genes we're inserting into crops are producing safe amounts of safe substances, there's nothing to fear from the genetic engineering process at all. The argument from artificiality is not a valid one without proof the additives produced are dangerous.
I disagree. The re-sequenced DNA modification is done to achieve a particular purpose, and it is tested and judged successful based on the results of that requirement. Side-effects must be obvious and relatively rapid in order to be considered as show-stoppers to release. We must accept that in the world of DNA, what we don't know about how genes express themselves in ways other than where we are looking is huge, vast. The bar should be far, far, higher than it currently is (which is virtually no bar at all, by the way).
There's also the worry of what happens when these artificial crops cross-breed, and on that topic I'm not knowledgable to comment.
The next generation of GM crops may be sterile in the second generation, which requires farmers to repurchase seed each year, and puts a stop to that fear.
However there are other issues. We are essentially working towards a single genetic expression of every given plant; in other words, a corn plant from a GM supplier is genetically identical to every other plant from that supplier. No diversity at all, which means that any disease or pest which adapts to eat or destroy that plant will be able to destroy them all. No hybridization means they are at a genetic dead-end evolutionarily speaking; they cannot mutate to adapt.
On a more commercial aspect, farmers are being sued - successfully - by GM crop makers for allowing GM seed to blow from one field onto another and germinate, even if it was not only unintentional, but despite the best efforts of the farmer to keep it from happening. Increased crop yields from GM crops ensure that farmers are essentially locked into a cycle of having to purchase what their neighbors do, or suffer lower yields and less money (high yields drive down prices, so they lose money if they don't keep up).
We also have issues of monopoly which are essentially left unaddressed and unrecognized in the US. Given current trends, there will shortly be only one or perhaps two producers of seed for all crops grown commercially in the US, and all of it will be identical genetically.
This seem like a good idea to anyone?
. Yes crops have been genetically modified over generations to favour the characteristics that we want but the abrupt and untestable-until-field GM process is one fraught with pitfalls that are not being addressed rigorously enough (given the consequences if it all goes wrong).
