I understand your concern, but it is usually less of a problem than it appears. In order for a claim to be held valid, or for a piece of prior art to be relevant to patentability, the invention must be enabled. An invention is enabled if the person skilled in the art would be capable of carrying out the invention as disclosed.
Here’s a hypothetical example. Jones develops a catalyst and reaction system that produces copolymers of ethylene with other olefins. These polymers have certain properties (molecular weight distribution, tacticity, etc.). The patent application provides actual examples of ethylene + propylene, ethylene + butylene, and ethylene + pentene. The patent also gives a laundry list of olefins, and vinyl-containing monomers in general that could be used in this reaction. Perhaps the whole list of copolymers with certain properties is claimed.
You go into the lab and successfully duplicate Jones work with ethylene + propylene / butylene / pentene. You then try to make polymers of ethylene + C6 to C10 olefins by Jones’ method and realize that it cannot be done. You modify the catalyst system and reaction conditions and develop polymers analogous to Jones’.
You file a patent application. In your specification you include examples of your invention and other examples showing that Jones’ method does not enable production of polymers of ethylene + C6 to C10 olefins. This information should allow you to defeat a lack of novelty (sec. 102) and hopefully an obviousness (sec. 103) rejection. (The lack of enablement in Jones’ patent would also be a defense against an infringement claim by Jones.)
Of course that’s not always the way it works in practice. A hurried examiner will typically see the Jones reference and give your application a rejection. Then it will be up to you to demonstrate that the rejection is improper.
Richard Tanzer