The United States uses a system called "peripheral claiming," in which patent applicants stake out the outermost boundaries or periphery of their claimed invention. This is in contrast to the older (think the 1800s) system of central claiming, where the applicant would claim the "heart" or "gist" of the invention. The problem with central claiming is that courts had to decide when an alleged infringer got "too close" to the claim, which was often a very subjective process. With peripheral claiming, the infringer either has each and every element of the claim or they don't (with a small amount of wiggle room for the doctrine of equivalents). This forces the applicant to say, up front, exactly what they've invented and—importantly—support it in the specification.
Over time, however, there has been a perceived (and perhaps real) growth in the number of patentees trying to make their claims cover more than they actually invented. This is what the written description and enablement requirements in patent law are designed to prevent. See 35 U.S.C. 112. I have said many times on Slashdot and elsewhere that these requirements should be tightened. What the EFF is proposing is, broadly, along those lines, though it is specific to software patents. I am generally not in favor of special rules for different areas of technology. Patent law has mostly avoided that. Copyright law has not, and it has turned the copyright statute into a sprawling mess dictated by special interests.
If the written description and enablement requirements were tightened up, however, then one way that computer-implemented inventions could demonstrate compliance (i.e. show that the application contained "a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains ... to make and use the same") would be to provide extensive source code or pseudocode examples. But in our zeal to rein in unsupported functional claims we should not return to central claiming.
As an aside, right now many if not most patents that involve software use flowcharts rather than pseudocode or source code. This is the result of Patent Office backwardness dating back to its initial opposition to allowing patents on computer-implemented inventions in the first place. Only now is the Patent Office finally recognizing that pseudocode or source code is much more searchable, readable, and useful for everyone involved, including patent examiners, competitors, and alleged infringers.