I am very fond of [XKCD](https://xkcd.com/), and that meme is pretty good too, [so's this](https://i.imgflip.com/tiec3.jpg). 

There really should be some screening & mitigation done at submission time _before_ block chain commit, right now that's occurring after the fact and the result is self evident. 

Such measures are never expected to be a 100% perfect robot nanny (deference to blind faith is misplaced anyway) because they would need to be somewhat lenient and permissive to reduce false positives becoming an obstacle, but it would help mitigate some blatant plagiarism and [tarpit](https://en.wikipedia.org/wiki/Tarpit_(networking)) submission of junk waffle posts while they are parsed & compared against criteria and again while the submitter considers editing and again if they choose to edit, consider:
- sircork spends 37 minutes writing an article then submits it
- system returns message of 13% match for `URL`
- sircork spends 2 minutes using markdown to indicate quoted portions of text
- sircork submits revised article
- article written to block chain

I can spend upwards of 5-7 hours on some of my articles that require writing/testing code, cross-browser compatibility checking, fact checking, collecting & verifying info/resources, capturing images of steps/results then editing them, and finally composing in markdown. I wouldn't care if I had to spend a few minutes more at submission time. But consider:
- user sees eye catching content & copies it
- user pastes content into steemit & submits
- user paid bot upvote army votes
- repeat `X` number times per hour every day

The article parsing will slow this, the system response (advice/warning) will slow this, even if they ignore it and hit `Continue` the article may be branded with the warning comment containing URLs of matching content & receive less or no reward.

The practice becomes less profitable and enticing, it may even become a nuisance to those who practice it, the quanta of these types of submissions is reduced and more of the reward pool remains.

As stated above, such measures aren't regarded as a 100% perfect magical digital genie that will fix everything and relieve humans of custodial duties or responsibility for actions, but rather a helper tool that does what it's able to, safely within a given level of certainty and reliability and no more, that stems the flood a little, making the task more achievable.