Yes this can be done and isn't a bad idea although it isn't perfect. My own philosophy is to oppose this format for a few reasons: How can the buyer know the invention is valuable without seeing it? (short answer: it can't be done) This would mean the seller would have to rely on a reputation system to sell his/her inventions at a decent price, with the price being only based on the quality of previous inventions sold. In my opinion, invention is one of the most egalitarian processes of our society, so relying on a purely hierarchical method of valuation (the reputation system) isn't ideal.
What you propose isn't far off from how many VC's operate today. They look for people with a reputation and history of business success and innovative ideas. This makes it almost impossible for people to have success with their first innovation no matter how good it is. Also, if you only invented one thing ever, and it happened to be the greatest invention of all time, you certainly wouldn't be compensated accordingly. Nobody is going to shell out 100 BTC to some "random" claiming to have reinvented the wheel.
The first message could be a very small fee, it would could be the proof, maybe a whitepaper, video, your own credentials, along with the addresses and price for the other sections of the invention.
As you get deeper into the invention the price goes up, until the final piece which at that point the whole thing can be created.
If you were doing this with bitcoin, the first message would be a video of a working QT, and the whitepaper, priced at a few pennies. The last message would be main.cpp, priced at thousands of dollars. The other modules would be in there, each at a little bit higher price. Yes at some point someone could get btc to work with their own main.cpp.
This is one approach. I think it is close to what I am proposing, but maybe not quite feasible with Bitcoin's block chain, because I've heard there are large fees for storing things in the blockchain.
Technical note: storing stuff in bitcoin's block chain would likely add more burden to the system and I doubt the fact that there is data hidden in the block chain with no utility to anybody else because it is encrypted will increase the overall value of bitcoin's market cap.
I like how you are thinking. For example: store your credentials as a movie maker, store a trailer for your upcoming film, sell the code/data for your hit film to the highest bidder. This is a fundamentally closed information system.
If a person buys the invention, he/she will likely keep the information undisclosed to prevent competition from making his new product, which is completely within his/her rights. Lets assume inventor1 and buyer1 have an agreement that inventor1 will not well sell or give away his invention. This means another inventor cannot come along and improve on the design without inventing the initial invention or owning the invention. Furthermore, there is incentive for a second inventor (inventor2) to invent THE SAME THING and sell the same invention through a completely different transaction to the same or a different buyer (a separate inventor spends his mental capacity reinventing the technology to sell it, after all the invention has never been publicly disclosed). This is basically a complete waste of society's mental resources which are finite. (two investments of brain power, one invention) That second inventor could have used his/her brain power to invent something completely novel.
I am striving for an open information system, like YouTube, but you get tradable credits for your submissions. Think back on the arguments for why bitcoin was valuable before people recognized it as valuable (8 months ago or more). One of the arguments was that it was more efficient/cheaper than other electronic payment systems, and as a result, the people who used bitcoin had more money to spend on things by saving from using bitcoin, so they naturally had more money to reinvest into BTC. That's how I want this economy to get its value, through increased efficiency. This increased efficiency was often pennies or even fractions of pennies per transaction. Keep in mind it is over $8000 to file a utility patent with the USPTO once, patents are often filed multiple times before being awarded, and each patent defense lawsuit can cost both sides millions of dollars in attorney fees. Nemesis will eliminate these inefficiencies.
You cannot create monetary value out of thin air. An argument for the value behind the USD is that it must be used to pay taxes. It must also be used to pay fees to the USPTO, which is essentially a tax. Citizens of the internet can take this value from government money for their own currency.