IOTA hacked or not?

IOTA vs MIT: whats the real deal?

Some days ago a report showed up, that Iota had a bug due to their created hashing function (CURL-P). The MIT (massachusets institute of technology) announced that they could create collisions in Iotas hashing function, which means they could create the same output with different inputs (they found with different data put into the hashing function). This should by definition not happen with a hashing function. After a week and a long discussion it seems that Iota knew about this. They say, that is was technically possible top hack Iota by reproducing their hashing, but practically it wasn't due to their trinary system and the one wayness of the hashing system. So who is right? 

The MIT claims, that the main problem with IOTA was, that they built their own cryptography...

Never build your own crypto
You might be a genius, but creating your own cryptography or hashing function might not end well, as you can't possibly see all the possible bugs other people might figure out. There are a lot of brains working differently, approaching differently, hacking differently. So better use a tested cryptography than trying to develop your own.

Iota's response:

You didn't hack us practically...
Iota in return says, that they knew their hashing function could create collisions. They say they needed to create their own hashing function (CURL) because of their ternary system (which is quite new). They knew they could get hacked, but with the coordinator the can secure the system, as it has the one-wayness. This means, the data put into the hashing function creates an output, but with the output only one can not reproduce or figure out the data being put in. Thats how collisions can be avoided. They say, a clone could be hacked because they wouldn't have the coordinator, but Iota itself can't. So Iota wasn't actually hacked, its system that was cloned (by MIT) was hacked instead.

What is this coordinator?

It basically is an assitant for the CURL, because tis new hashing function is new and in its enfant stage. Or as David Sonstebo (the founder of IOTA) says:

The coordinator is like training wheels for the network until the amount of organic activity on the ledger is sufficient to where it can evolve unassisted

So my question is: If the coordinator provides you to figure out the input, but MIT figured out different inputs to create the same output, it still is able to be hacked? I'm new, maybe you understand this and can help me?? Iota itself for me is still a very interesting project. Here the video for more information, this guy seems to know what he is talking about.

For more informaiton about IOTA:

https://blog.iota.org/the-transparency-compendium-26aa5bb8e260

CHeers