This is problematic - you can imagine a NPE taking this to heart and simply producing proteins for all the known human genes and contesting any future treatments based on protein therapy.
Perhaps recombinant gene therapies should be regulated in the same way as pharmaceuticals where the first company to market would be protected for some time from competitors who would sell the same compound. In this system, using an unrelated peptide that had the same effect as insulin would be permissible but simply re-cloning human insulin would not be.
As a side-note, I think there was a technically-innovative approach [1] used to make humulin which involved two non-naturally occuring nucleotide sequences. I suppose these might be patentable for glabrifons'...
1 - Unfortunately, I can only find this source: http://books.google.com/books?id=gP4V4uGqp94C&pg=PA755&#...
However, if someone were to use the same process they used for gene insertion (which is more likely to be an invention than a gene they simply discovered), they may have grounds to sue.
The process for gene insertion for insulin was the Herbert/Boyer recombinant DNA patent that was widely licensed by Stanford: http://www.genome.duke.edu/centers/cpg/case-histories/semina...
If I had to pick Genentech protecting their drug via a patent on an E.coli bacteria containing human insulin gene vs. the method to move the gene I'd take the former for sure. The method was much more broadly useful.
Monsanto is just being greedy here, and the net impact on their profits will be minimal even if they lose. There will always be a market for them to sell engineered life "factory fresh" even if people use secondary generation of seed.
The first major biotech "invention" was Humulin. Basically putting the human gene for insulin in bacteria to produce human insulin. Humulin launched the biotech industry. You should consider how your logic applies to that case. While I find your reasoning appealing, I don't think it is sound or applicable.