Perhaps have a think about other engines, some have huge scope for improvement, others only a little.
Just quickly -
A naturally aspirated petrol car engine can usually gain only 5-10% with simple tuning mods, but even doing this will require pushing the timing closer to detonation and you will need to be fussy about the fuel that goes in it.
A turbo diesel motor can usually gain 30%, but push it hard and hot and the exhaust valves burn up.
Turbo petrol, gains vary greatly but 10%-30% is typical and easy, but you must live with the detonation risks as you wind up the boost.
Bigger gains than this are usually just moving the power curve higher in the rev range and require major changes to engines and supporting systems.
In some of these cases drivability would suffer, more top end power at the expense of low end torque for example.
Noise and pollution rises - both things that are regulated for a manufacturer.
Now if you look at mega buck cars, well, I cant imagine you can turn the wick up on a bugatti veyron and expect much more. And thats the cost point, chainsaws are cheap.
The prices that modifiers here charge - if added to the wholesale cost of a saw would double the price of a new saw. And that would just cover the "blueprinting" aspect cleaning up the castings, setting squish, port matching, etc. This MIGHT gain another 10%, would you pay twice as much for a saw that got 10% extra power? Without muffler mods and staying within exhaust emission limits I doubt that any more than that could be likely. And you'd have a saw that could ONLY run on decent quality fuel, stale low octane stuff would be a recipe for disaster.
I'm just talking basic physics and manufacturing realities, experienced tuners would no doubt have specific examples that contradict what I've said, but as a wild generalisation this is true across all examples of the internal combustion engine I think its fair.