Steel is strong, but its specific strength is less than either aluminium or titanium. The reason you can build a steel mtb frame with thinner walls than either aluminium or titanium is simply that its density is greater.
The tube you quoted, 31.8 x 9-6-9, would be a reasonably substantial top tube, but you won't often see a standard 4130 top tube of less than 28.6 x 9-6-9. Heat treating makes the material stronger, so an mtb can be made with heat-treated steel with much thinner tubes. The Columbus Cyber top tube on my Kilauea is a 28.6 x 7-4-7, which is about as thin as it gets for mtb.
Aluminium tubes aren't all that much thicker. Easton Ultralite has a 0.8mm top tube, and Easton 7000 Scandium went down to 0.65. In titanium, you tend to see 0.9mm plain gauge as the thinnest. It has the opposite limiting factor to aluminium - it could go to similar thicknesses without breaking, but butted titanium isn't often used for mtb because it would be too flexy. The limiting factor for aluminium is breakage - it is still quite stiff at 0.8mm.
I don't think there's any justification for thinking in terms of 'scary light'. Frame builders operate within the terms of the warranty given by the tube manufacturer. The manufacturer will specify the thinnest selection of tubes that are suitable for mtb and no frame builder will go below that for a production frame.
Aside from the warranty costs, there's nothing to stop a builder using road tubes for an mtb frame, but the only case that I've heard of was mid 90s Ritchey team frames, some of which are lighter than an mtb tubeset. Even there though, Tom Ritchey knows a lot about tubes and would have taken into account (a) that Tange were quite conservative with their specification (8-5-8 being their lightest for mtb) and (b) that a brazed join weakens the tube less than a TIG weld does. He used to say those frames were built to last no more than one racing season, and some of them didn't even make it that far. Whether they were actually faster than a slightly heavier but stiffer frame is another issue. Only for a very light rider I would guess.
The tube you quoted, 31.8 x 9-6-9, would be a reasonably substantial top tube, but you won't often see a standard 4130 top tube of less than 28.6 x 9-6-9. Heat treating makes the material stronger, so an mtb can be made with heat-treated steel with much thinner tubes. The Columbus Cyber top tube on my Kilauea is a 28.6 x 7-4-7, which is about as thin as it gets for mtb.
Aluminium tubes aren't all that much thicker. Easton Ultralite has a 0.8mm top tube, and Easton 7000 Scandium went down to 0.65. In titanium, you tend to see 0.9mm plain gauge as the thinnest. It has the opposite limiting factor to aluminium - it could go to similar thicknesses without breaking, but butted titanium isn't often used for mtb because it would be too flexy. The limiting factor for aluminium is breakage - it is still quite stiff at 0.8mm.
I don't think there's any justification for thinking in terms of 'scary light'. Frame builders operate within the terms of the warranty given by the tube manufacturer. The manufacturer will specify the thinnest selection of tubes that are suitable for mtb and no frame builder will go below that for a production frame.
Aside from the warranty costs, there's nothing to stop a builder using road tubes for an mtb frame, but the only case that I've heard of was mid 90s Ritchey team frames, some of which are lighter than an mtb tubeset. Even there though, Tom Ritchey knows a lot about tubes and would have taken into account (a) that Tange were quite conservative with their specification (8-5-8 being their lightest for mtb) and (b) that a brazed join weakens the tube less than a TIG weld does. He used to say those frames were built to last no more than one racing season, and some of them didn't even make it that far. Whether they were actually faster than a slightly heavier but stiffer frame is another issue. Only for a very light rider I would guess.
