Besides what Wagi already said, the one thing that really got to me was the complete lack of character identity.
The problem with Halo machinima is that nobody's mouth moves, and when you've got 4 characters on screen, it's nigh impossible to tell which one is supposedly producing the voice. This combined with the total lack of individuality to the characters (there was the sweary posh one, the sweary whispery slightly asian one, the sweary one and the other sweary one) I could never tell which voice was who.
The swearing, I think, would have been much more appropriate and possibly even quite funny if it had all been confined to just one character (though without such copious use of the N-word in particular) rather than every character conversing through swears alone.
For future projects, keep just one character on screen when they are speaking, at least for the first time they do speak, so we can tell which voice is which character. Also, stay out of Team Deatmatches, because everyone suddenly turning red and blue instead of their normal colours was a hell of a way to throw away any previously established character recognition.
Also, don't be afraid to vary the voice acting more wildly, draft in some people of different nationalities and ages, as the range of voices you can use increases, more characters can be introduced.
TL;DR - What Wagi said + only (maximum) one sweary bloke, vary voices and characteristics more, one person on screen at a time when they speak.