Well, as after the interbreeding the survivors have 1-2% of Neanderthals and 98-99% of humans, and none with the opposite proportion have survived, that counts as outcompeting and very one-sided assimilation.
Similarly, saying modern humans have 98% modern human DNA seems a bit circular. And how is that 98% calculated? Don't we share 98% DNA with chimpanzees and bananas or whatever? Is it 99% of that final 1%?
A more relevant example: the diversity of genetics in Africa could be counted as like 20 "races" if races were based on genetics, but instead we generally lump them mostly into 1 of 4-5 vague historical socioeconomic groupings. So you have one "white" parent and you're mixed race, but you could have two more genetically different African parents and not be considered mixed. It's nostly about where you draw subjective lines.
When we say we have 2% Neanderthal DNA (in Europe, less elsewhere), this is usually done with SNPs [0] which are used as a way to quickly differentiate between populations without sequencing the whole genome. In practice when companies like 23 and me tell you that you are 20% Pacific Islander it means that they have detected a number of single base-pair variations that is statistically compatible with that result, but it doesn't mean that Pacific Islanders and Scandinavians are not largely the same genetically, just that we know some easy places to look to tell them apart.
I should probably also mention that one interesting and politically relevant result of these population studies is that the concept of race doesn't really have a solid statistical base, because we have found the variation inside the group to be so large compared to the variation across groups, that there is no unambiguous way to cluster the human population.
[0] https://en.wikipedia.org/wiki/Single-nucleotide_polymorphism
If you concentrate on a large pool of genes, yes. But then you might come to the same findings when studying dogs or cats and conclude that breeds don't have a solid statistical base.
With statistics everything is provable, given carefully selection of samples and traits.
What would be a differentiator - aka different brain-chemistrys and neuro-types - are non-existent or omnipresent.
Taking neuro-types as adapted to circumstances, the characteristika usually fielded by racists, are almost always a expression of "percentages" of neuro-types thriving due to different historic circumstances.
If you spend several generations enslaved and waged war upon, those neuro-typus adapted to the circumstances - will make up a large part of the surviving population.
Racism causes history, and history causes perceivable "Race" because humanity adapts to whatever it goes through. Given enough "good time" generations, the "adapted survivors" retreat percentage wise and are replaced once again with peaceful neuro-types. Unless, the stressful conditions are artificially prolonged.
Consider Neanderthals had more sparse populations due to being adapted to colder climates.
In periods over the millenia, mixed groups could have lived without being aware of being mixed at all. Groups with 50% heritage or more could have been small compared to the growth of modern human populations elsewhere, which would eventually blend.
Over 100 generations, this 2% heritage could be accounted for by a growth-rate difference of only around 5% vs 9-10%, which is just above replacement level and this difference wouldn't have been noticeable to each generation.
If one population had growth spurts and doubled over a generation, then much faster.
Also, population growth also counts as outcompeting. If after thousands of years there’s only Homo sapiens and no Neanderthals, then that makes the point. Homo sapien genes won out.