I don't think the effect of VA on sales of realAcorn hardware is as simple as some people make out. Sure there will be some people who bought VA and chose not to buy new the realAcorn hardware that they could have bought. Mainly these will be people who might have bought an Iyonix. However, I guess for many of them buying realAcorn wasn't a proposition because they needed a portable machne (which remains non-existent) or because they also needed a realWindows machine because of the limitations of RISC OS.
What strikes me now is that the people who bought VA because realAcorn wasn't a proposition may now be changing their minds with the advent of a cheaper realAcorn machine. So in these cases VA has kept them using RISC OS when they could easily have gone over completely to another OS - and now they might well be buying new real Acorn hardware. If this is the case, and I suspect there is something in it, then VA will have helped sales of realAcorn by keeping them in the game.
Obviously it's Martin's call but I think it may be time for a different sales model to help Artworks development. I bet there are lots of RISC OS users out there looking rather enviously at what Artworks can do but not prepared to pay the not so insignificant sum required. RiScript may show the way.
Market a 'lite' version of Artworks for a paltry sum (less than 50 quid) but without the latest modules. The sell the extra modules at a price which should be able to subsidise more development. I can't justify 150 quid all in one go - but I could envisage spending something approaching this spread over a period. Furthermore, I bet there are quite a few like me.