Qdog, thanks for the picture -- I think I see what you have in mind. Some of the key goals that I am trying to achieve as I think about refining the design are 1) ease of use, including choice of curd size; 2) ease of cleaning -- ideally, every part that contacts food, even indirectly, can be tossed in the dishwasher; 3) ease of manufacturing; 4) cost -- partly a function of required materials, but even more a function of #3. My current design satisfies most of these very well; where it falls down a bit is in #1 -- it is easy to set up for the desired indexing, and the indexing is precise, but it is rather stiff to move it out of one index to the next.
It seems to me that your suggested design would have the advantage of being easy to move from position to position, and easy to release the rod to move it by pressing on the lever. A couple of disadvantages: all of the indexing would be in the rod (i.e., it would need to be grooved every 1/4"), and thus there would be no way to select a different indexing (e.g., 1/2" or 1") without either changing to a different rod or having to count (e.g., counting 4 indexes each time for 1" curds.) A bigger disadvantage, though, would be that this design would be a good bit more complicated to manufacture, and if it used a spring, a good bit harder to clean.
I had originally thought about, and experimented with, a different sort of spring-driven design, using a spring-loaded ball to engage the slots in the rod ... but my prototype did not perform very well, and it became obvious in the process that it would be a good bit of work to make, and a royal pain to clean. Then I thought of the current design -- much easier to make and to clean, but as I said, stiff to move out of one index to the next. I had been hoping,when I got time to experiment further, that I could fine-tune either the clips or the depth of the slots in the hub to get the tension just right, so that it is easier to slide from index to index while still being securely indexed.
However, one part of your suggested design might just be the key to improvement. Currently the indexing slots in the rod of my design are shaped something like this: \__/. I kept the slope on the sides relatively steep to ensure security of the indexing (particularly when I was thinking of engaging the slots with a spring-loaded ball), but that is a large part of what makes it stiff to pull out of one index to move to the next. But what if I cut the slots in the rod to the shape shown in your sketch -- something with a shallow slope on one side, but a 90˚ angle on the other side. That would make the rod able to index only in one direction, but it would be much easier to pull it up to the next index; meanwhile, the 90˚ side would provide a firm index, so long as one kept just a bit of downward pressure, which should happen naturally in the process of using it. I will definitely give this a try ...
Keep the ideas and suggestions coming -- maybe between all of us, we can come up with the perfect design!