There has been a great deal of effort put forward over millennia to answer the Problem of Evil. the Problem of Evil, however, is still generally considered a "problem", and has been for most of those millennia. Given this, it is generally good practice to presume that a particular answer is likely not sufficient in one way or another.
In trying to articulate and answer the Problem of Evil, it is important to clarify some starting positions.
First, while Scholatics, notably Aquinas, attempted to show that a God must of necessity have some characteristics, these efforts have generally been considered unsuccessful. One can, for instance, seemingly quite reasonably postulate a non-omnipotent God, or one with moral imperfections. If these were necessary conditions, then the logical flaw in such postulations would be readily demonstrable.
Therefore when one proposes a God of X characteristics, THIS IS A POSTULATION, IE a hypothesis. And the Problem of Evil, is a TEST of the hypothesis that God is Omni-benevolent.
A further point is that reasoning and logic sometimes break down around infinities and zero. Examples are divide by zero and divide by infinity operations. 0/0 is particularly problematic. To avoid zero/infinity logic glitches, it is better to discuss Omni-properties in less absolutest terms. Omnipotence need not be infinite -- it need only be sufficient power to restructure anything there is in the universe. Omniscience need not be infinite, just an awareness of everything in the universe (and past, and maybe future). Omni-benevolence -- is harder to constrain by the universe -- one can instead go with -- recognizably more moral than anything we can imagine we humans coming up with on our own.
Now, when one starts with multiple postulates, it is in the nature of derived systems that they end up with contradictions. Omni-benevolence COULD impinge on Omnipotence -- that is why legal systems so often end up in tangled contradictory knots, and judicial systems are continually working to resolve these contradictions. Trying to define such contradictions away -- is generally doomed as each postulated non-contradiction principle ends up multiplying the postulates, and would therefore be likely to lead to yet more contradictions.
In pragmatic terms, contradictions are indeterminate cases -- where the system could arrive at two or more different states, based on the starting premises. Could an omnipotent God make some matter so robust even He cannot then destroy it? Maybe. By the initial definition of omnipotence, no, but then the ability to modify the base nature of matter is included, so yes, but then God could modify it back, so no, or God could forget how to re-modify, so yes, etc. One can simply accept that indeterminate cases are -- indeterminate.
Your approach to "logic" is also questionable, and brings up another issue. You refer to God as a creator. This needs to be specified explicitly, not just mentioned as an aside, and it is another major postulate. The term often used in early Christian writing is PantoKrator -- creator of everything. If you go with PantoKrator, then you will have to figure out how to deal with logic. Did God create logic? If not, then where did it come from? The answer could lead to the pasting in of more postulates ...
At any rate, you have correctly realized that a morally imperfect world is possible, hence a PantoKrator is not necessarily omni-benevolent. But no, this does not mean that your postulated God is not "essentially" good. I think you are confusing your hypothesis, with a necessity claim. PantoKrator is one postulate. Omnipotence is another. Omni-benevolence, IE that God is "essentially" good, is just a third postulate of your hypothesis.