In theory you're right. In practice you're wrong.

What you say is true only under perfect conditions (like looking for a non-rating black hole). In practice much of a program's speed comes from the algorithms and code organization instead of small-things optimization. For example, in theory it would be faster if you used the assembly language but in practice this will result in slower code since you'll not implement common patterns as good as they are already implemented by any other language (or its libraries).

That's why it may be faster to write something in python using Qt bindings (or java) than write it in C++, since you'll take advantage of some very efficient data handling methods that you'll most probably won't implement yourself.

Of course there are many exception to the above (e.g. linux kernel), but when dealing with common program developers like ourselves, the above becomes mostly true.

However, there are other benefits in native code than the speed and that's where you should focus instead.