Having now listened through some Numenera actual play, I noticed a couple of things that felt worth mentioning as an addendum.
The Roo Sack Gamers used an awful lot more Effort, Cyphers and other modifiers than we tended to, and this lead to a much more variable set of target numbers, as well as changing the balance of fights. While we were quite cagey about Effort, and used it about twice, they took it as a core mechanic pretty much straight away. In some cases this allowed them to crush enemies savagely, but in others it backfired due to poor rolls and/or overEfforting. If you’re expending Effort to adjust your success rate, both in combat and outside it, that effectively causes hit point loss, which means fights that might have been simple otherwise become a lot more challenging. I feel like that somewhat addresses Dan’s issue of the fixed target numbers making things bland, as well as our sense that combat wasn’t very difficult. It doesn't, however, address Arthur's concern that spending Effort always harms you while only occasionally helping.
Arthur also suggested Numenera feels like one of those games that can only really be run as intended by the designer, due to their very idiosyncratic setting. Certainly, I felt that Numenera didn't convey a single strong vision to me and give me a sense of what the world feels like. However, Roy seemed to have got a solid grasp on something, whether or not it was what Monte Cook had in mind, and I found the presentation as convincing as most other fantastical settings tend to be. He was fairly easily able to improvise largish chunks of content without me being able to tell what was planned and what was improvised - in fact, I assumed he might be running or adapting a published adventure, but having asked him, apparently not.
The assaxin problem remains, and the subtractive armour system looks like it is going to throw up occasional issues. In the RSG campaign, there was one fight where a character completely shrugged off all his hits because of heavy armour. Very bad defence rolls can still increase the damage to break through armour, and one option would certainly be for the GM to adapt by having intelligent enemies switch to other tactics when they realise armour’s too thick – trying to knock down or disarm a PC, for example, or grapple them and expose the PC’s weak points to another enemy. I suspect the main issue is going to be, as with the diablodon, when PCs find themselves unable to inflict damage on enemies. With light weapons doing only 2 damage, it’s relatively easy (as far as I can tell) for a non-Glaive PC to find themselves needing a 17+ to get bonus damage in order to inflict any injury at all. Again, perhaps at this point they need to start playing cunningly and finding other ways to defeat enemies, but it could easily get just plain annoying, at least if the combat is hard to avoid. There's only so many combats based around setting up indirect tactics you want at a time.