The problem with proposals for political reform is that the subject isn't very exciting. Proposals could also be quite technical and these things could make it difficult to get a referendum on a proposal if the signature barrier is set too high. The other side of that problem is that if this barrier is set too low the system will be swamped with impractical proposals. The answer to this problem is to set the signature barrier low, perhaps around ten thousand, but have a filter to take out any unrealistic proposals.
Our filter will be a commission of twelve people selected by a combination of lot and election. Our suggestion is that at every election two people are selected by lot from each constituency to give us a panel of eighty. Some will be willing to do this work and others will not. They will select twelve people from the eighty to be commissioners for two and a half years, and when that period is up select another twelve for the remainder of the term. This will give us a truly independent commission, but they're not going to be busy. Few poor proposals will get the required number of signatures and good proposals will be adopted without going to the commission. If you have a reform proposal you don't start by collecting signatures, you take your proposal to your elected representatives. If they reject good proposals and force people to go to the effort and expense of collecting signatures and running referendums unnecessarily it will cost them votes at elections. This also applies to proposals for ordinary legislation because voters will punish parties that refuse to support good initiatives.