The other definitions are not redefinitions of racism, but appropriations of the more general definition to further the term. All depend on the notion that racism is dividing people into groups called races which contain people with identifiably similar traits.
As someone who has spent a significant portion of life in a humanities department, I'll say it's a stretch to say 'most' academic work goes by the vague definition posted above. Frankly that definition sounds like something I'd encounter in a first year paper rather than something a colleague would agree with, much less endorse. Further the academic world is much bigger than sociology, and even within the social sciences (e.g. economics) there is an insistence on using less fluffy and more logically consistent definitions.
A definition that makes sense and has no meaning. I don't know about racism, but you certainly seem to have lost sight of what "meaning" means.
I've never had any trouble understanding this definition of racism. When I was young and stupid, I thought myself extremely clever for seeing instances of reverse racism, but this was because I understood nothing about racism except that MLK Jr. gave this nice speech a long time ago and now bad things no longer happened. I thought it was this magical thing where there were black people and there were white people and I was yellow except not really so it was kinda weird and I didn't understand why the red people didn't seem very red.
Then I grew up and started understanding how power moves and manifests. It's very fuzzy. You don't see mathematical equations about how the election of the POTUS changes opinion in the Middle East, because we haven't figured out how to model that. Our understanding of power is extremely weak, compared to something like how many atoms of hydrogen are found in a molecule of water. To make this extremely stark, we don't understand power. We have a feel. An intuitive notion. That is all we have.
We do not have explicit forms of racial discrimination to any significant degree. But we do still have significant power imbalances that map suspiciously well to racial divides. These power imbalances are virtually impossible to quantify, because we have no idea how to do it, but we can infer them from statistical trends. We've chosen to call this racism.
That's language drift for you. There are reasons not to call it racism. Apologists, such as yourself, have enumerated a good number of them for us. But there are also good reasons to call it racism. First among these is that we do not need the more generic definition any longer. English speakers generally have difficulty finding instances of chattel slavery or explicit segregation laws. In both of these cases, we have more specific terms anyways. Second is that it signals the correct emotional reaction. Most people are offended when called a racist, because they've been trained to understand it is a bad thing to be. Used properly, it forms a foundation to change behavior. (And before you say that it is sometimes used improperly, this would be true of a different term as well; but a different term would not have the same, useful emotional charge.) Third is that the issue remains one of "dividing people into groups called races which contain people with identifiably similar traits". The generic definition still actually applies, but it does so more weakly.
I'm sure you take issue with all of these, but really? The definition could be improved. That doesn't make it meaningless, as demonstrated by the many people who use it in that capacity without a problem. It does make it difficult, as demonstrated by the many people who seem incapable of understanding it.
Conflating the two definitions and describing people as 'racist' for re-enforcing racial imbalances in society is disingenuous.
A lot of people oppose affirmative action (thus earning the label 'racist' according to the latter definition) not because they are prejudiced, but simply because they do not believe in the efficacy of the process.
As with many things in life, it's not necessarily what you say, but how you say it that determines context and intent.
Policies that acknowledge and address systemic racism are not "racist" because they acknowledge race and make decisions based on these imbalances.
This is not my definition, there are loads of people with this definition.
Rational discussion can only occur once terms have been agreed and defined
Agreed, that's why I made this point. Too often people are only aware of the first definition.