is it though? this is or was available worldwide, and not every country has a history of blackface.
i think it would have been way more racist to exclude African features on the assumption that black people are too fragile to handle the app, tbh.. thats just my opinion though and im not black so fuck if i know.
i will say though, we can be pretty sure that the app makers did not take into account all the different facial features in Africa, because in reality there is of course tremendous variety (i would guess at least as much as in Europe or Asia, probably more) .. Somalis, Nigerians, South Africans, Masai people etc etc all look very different thanks to the huge genetic variety on the continent. so maybe its racist because they simplified it too much.. i can buy that.
Or if the app was only usable by white people, as the point is anyone anywhere in the world can enter their picture and see what the tech can do.
The way some may have used it online is what could've been causing some issues, but like many tools people online can use them in dubious ways. Google fan art for certain games and then find out your eyes and sensibilities come under attack. People have been photoshopping and editing things for a long time, and they key thing to remember here is that the app was an equal opportunity provider. Usable by all for all.
I think in 10-15 years time something like this releasing wouldn't bat an eyelid. Right now it's probably an example of leading tech in its field given it's free and can run on a mobile phone, and is going through that slightly hyperbolic public backlash as people aren't used to this kind of ease of tech dramatically and realistically changing their appearance. It makes many feel uncomfortable, but in reality, as a social experiment, it's probably a good thing if that uncomfortability is challenged as why are you feeling that way about other racial differences, skin colors, etc? Pie in the sky wishes as an app isn't going to radically alter perceptions, stigma and so on but it could maybe do a little in the greater battle.
Give it a bit of time and the tech will be even more advanced, impressive and probably being used in another ways. I could maybe even see something like this in the future expand into being used in games development for quick and accurate diversifying of character models. Look at how decently it can work already on 3D models and even games. That FF13 picture is quite impressive. Yes it needs more diversity within the filters, but it's arguably fledgling tech, or tech that is just getting started and will no doubt be refined and added to along the way.