Red Eclipse on the tesseract engine may be very different from Tess, but I can picture it being a lot more demanding than even Tess currently is
this old PC i play on at the moment (core 2 duo, Radeon HD 5450, 4GB RAM) can manage nice playable frame rates on semi-demanding games at medium settings (crysis, skyrim, bioshock), but i get no more than 25fps on low-mid settings in Tesseract, even on the simple deathmatch maps in the first release (in complex it's much worse). this is without any other players/bots or action happening. Tess is very demanding and poorly optimized, it's free after all and not made by a professional studio with a team of developers. it looks very beautiful in practice but a lot of that is due to the high amount of map-models and high-res textures. RE would be ported with what it's currently got.
my setup or x setup is no way to measure true performance, some thorough tests should be done to see if the majority of the existing RE playerbase will struggle to achieve playable frame-rates on this new engine. and if it turns out that's the case, then it's a bad idea, regardless of the very obvious improvements it offers. RE could stand to improve a lot before a drastic change like this anyway, perhaps it's not done with it's journey in the Cube 2 engine.
conversely, the improvements it offers could
potentially attract new players with it's nicer aesthetics, the breed of gamer with powerful machines that could easily max out the new visual implementations. the ultimate question is whether or not it's worth the risk (Tesseract doesn't have a playerbase worth speaking of at the moment)
additionally i don't think this decision should hinder the rush to a 1.5 release, i mean come on already