That was not my point. Ofc we can improve ISP. No idea how much that improves either 'performance' or drops price.
It improves performance a *lot*. As for price, it depends on how expensive that rocket system is. For first stages, an improvement in ISP's effect on the size of the rocket isn't that much greater than linear. But the further up the delta-V chain the engine is used, the more of an impact it has on everything that was used to get it there. An extra hundred sec ISP on a first stage might reduce the system mass by a third; on a second stage up to LEO, maybe cut it in half; on a kick stage for a Mars transfer orbit, maybe cut it by two thirds. On an ascent stage from the surface of Mars... well you get the idea. Shrinking down a rocket to a small fraction of its size - fuel, tankage, and engines - well, that's really significant. ISP is very, very important for upper stages. So you can afford to pay quite a bit for those top stages if it improves their performance. Just not an "unlimited" amount.
There is no way a high tech electrical engine will improve its performance by 10% regardless how much money or time you put into it: the efficiency is already between 98.5% - 99.5%, up to 99.9% in some cases.
This is getting a bit offtopic, but at least the electric engines in EVs don't usually run at nearly that high. Depending on the type they might average 85 to 94% on average. It varies over their load cycle.
Regarding rockets: there is simply not much margin anymore in changing the form of the exhaust tube, burn chamber etc
Actually you can. The general principles of how rocket engines work are fixed, of course - your exhaust will never exceed its local speed of sound in the throat, and then you want to expand it as close to ambient pressure as you want. But the details vary greatly. There's bell nozzles, linear nozzles, annular nozzles, aerospikes, throatless nozzles, atmospheric wake compression, and on and on. There's tons of different ways - developed, in development, and in theory - to pump and inject your propellants - where they need to be pumped at all. Even many propellants that are traditionally thought of as being in one state can be implemented in other states. There's various ways - developed, in development, and in theory - to prevent nozzle erosion. To improve regeneration. To reduce mass. And on and on and on. Rocket combustion is a rather complex thing and we're still trying to get a handle on it. Do you know that we still really don't know how aluminum burns in solid rocket propellant? There's something like five different competing theories. I mean, things like this are a Big Freaking Deal(TM), especially when such small improvements in upper stage ISP have such significance for lower stage mass. And even on your lower stages there's a lot of things that have a big effect on your system cost. For example, how to stop resonant shocks from ripping them up - a lot of people don't realize that one of the main benefits of adding aluminum first stage to propellant mixes is that the droplets of burning aluminum damp shocks. (yeah, it increases ISP too by raising the exhaust temperature, but it also has disadvantages, such as not contributing to expansion, slowing down gases (particularly near the nozzle), and impacting/eroding the throat (or even forming an accumulating slag)
Re, nuclear+chemical. There are proposals for this. The main issue isn't efficiency - the extra chemical energy doesn't make that much of a difference - but thrust. The downside to nuclear thermal is that the reactor is so heavy (fission is like that, unfortunately) that the mass ratio is only something like 3-4:1. That's really bad (you generally get 15-20:1 or even better for a chemical first stage). So the approach is to inject oxygen early in the ascent phase for added thrust, but only run on hydrogen higher up when gravity losses are lower. I'm really not that sanguine about nuclear thermal rockets getting a serious development program any time soon, though. The public overestimates the risk, of course - not only am I sure they'd well seal the fuel elements against whatever damage would be incurred by explosion or reentry, but there's the simple fact that the fuel is "fresh", not contaminated with the more hazardous actinides. But it's going to be a hard sell. And a really hard development project, if they ever did try again. Gigawatt-scale flying nuclear reactors that pose radiation hazards during assembly and test aren't exactly childs' play.