I think it would help to first focus on "part of speech". You are relying on traditional philosophical categorization (elementary school grammar slogans like "person, place or thing", so I guess you surmise that "creates" is a verb because it isn't a person, place or thing. You guessed right that "creates" involves an inflectional suffix, but is that really self-evident, or is that just a subconscious and in this instance correct calculation – based on what?
Part of that process involves classifying words in terms of part of speech, which we do not by intuiting meaning properties, rather we look at distribution in sentences. This is the classic frame analysis done in intro syntax, i.e. "He ___ a car", "They ___ a car"... We identify nouns, verbs and adjectives based on where they can appear in various sentential slots.
One property taken to define the distinction between "inflection" vs. "derivation" is that adding an inflectional morpheme does not change change part of speech. Accordingly, there are at least two homophonous affixes -ed, one for adjectives (you can call it something else if you want) as in "a created paradox" versus a verb in the past ("he created a paradox"). Everybody knows that there are multiple -s morphemes, which happen to be inflectional.
Another criterion relevant to inflection vs. derivation is murkier, namely whether the function of the affix is "more syntactic" versus "more semantic". The considerations that lead you to think "more syntactic" are basically sentential co-occurrence, for example "He create-s a mess", "I create-Ø a mess", but *"He create-Ø a mess", *"I create-s a mess".
The suffix -ing might therefore be called inflection, because it appears in a syntactically-defined frame (copula __ verb-whatever) as in "He is creating a mess", so it could be an instance of a general verbal inflectional concept "tense-aspect", but then you can also say "I hate his singing", and not *"I hate his sings", so it isn't always a verb inflection. Therefore, there are at least two affixes -ing.
Once you step outside English and/or Indo-European, the distinction becomes even less tenable, and doesn't play a particular big role in grammatical theory. What does seem to matter is "agreement", that a verb may have to "agree" with other words in certain properties, which is a fundamental syntactic fact that a grammar should capture.