In case I may have made finding them sound easier than it really is - to find something like this on ebay, you have to set a saved search for something like butcher chisels or another early sheffield name, an attribute and sit and wait for ebay to feed you what's new each day. 

You end up with a harlequin set of chisels, and not a matched set. there are a lot of very short chisels left from this era, and others that have been setup in a way that's not going to be easy to recover or with very deep pitting. 

To get a sense for my early comment about sandstones, someone has to try something like a later ward that matches well to a very crisp finished edge off of something like a washita vs. an all sandstone or sandstone and slate regimen. the same dynamic exists with japanese chisels - the older chisels and plane irons that were tempered too hard to work with natural finish stones are often poorly prepared or little used. 

I don't do much carving, so I'll yield to carvers on what's better, but addis tools had a reputation for being undertempered. I wouldn't like a gouge that chips too easily, but one that's just a bit on the softer side is quite pleasant, and the same is true for complex moulding irons and H&Rs. there's less to gain there chasing hardness. vs what you can get out of a smoothing plane or more generalized set of chisels. To chase higher hardness, everything has to be right with alloying and carbon, and then the stones need to be able to handle it. 26c3 is super easy to sharpen, but the hardness would make it real trouble for sandstone and slate. India followed by washita and it's biting and general sharpening stays ahead of damage. 

The last comment I think would have a whole lot of historical use just by sense of a user - that a tool has to stay undamaged in general use and whatever damage occurs has to be removed as a matter of routine sharpening. As stones become marginal for a hardness level, staying ahead of typical day to day nits becomes harder. 

this is a matter of speculation based on economics, not ideals. it's the same reasoning about why the double iron eliminated single iron planes for anything where volume of wood planed starts to count and the surface is visible. They are a significant economic advantage. Once a machine planer and/or drum sander gets involved, then, they aren't. In perusing guitar forums, I've seen hand planes mentioned other than violin planes (for carving tops) exactly zero times. 

getting a better result or a good result faster always wins.