also known as a k of n multisig

this means that in a 2-of-5 multisig where 3 people dont have to sign,

as oppose to

interactive, where by its n of n, meaning it has to be a 5 of 5 or a 3 of 3 where everyone has to sign

No, that is not what interactive means.

In traditional cryptography, interactive means that the participants in the multisig must verify that the other participants actually control the private keys to their public keys, not that it is n-of-n. Non-interactive does not mean that it is k-of-n, it just means that there is no key exchange and verification protocol that must be executed.

Anyways, MuSig is specifically an n-of-n multisig scheme as it is a multisig scheme in the context of traditional cryptography.

this means when users see the funding public key, they dont know how many other signers there are or are needed in total

The same is with current multisig using P2SH. Hell, you don't even know whether a given P2SH address is multisig or not, so if you are a participant in any multisig, you should always be asking for the redeemScript, or, if using something based on MuSig, you should be asking for the public keys involved and make sure they calculate to the given aggregate public key.

and when the funds do move.. never get to know who specifically did sign

because the address doesnt tell them it is a 2 of 5..

because the address doesnt tell them who the other 4 people are

making it easy for whomever set it up to tell 3 guys its a 3 of 3 when in reality its a 2 of 5, and in reality whomever set it up owns 2 keys himself

so the 3 innocent guys dont realise its a 5 counterparty address.. all they can see is that their key is part of AN address.. but not know how much of a part...

This is completely incorrect, did you even read the paper or the linked blog post?

As I said earlier, MuSig is

**only** a n-of-n scheme so this concern is completely invalid. To do k-of-n, additional work is required.

I don't quite get yet how the k-of-n threshold scheme comes in (maybe it's out of the scope of that paper?), but in general, this is great news. I hope a BIP detailing the actual implementation into Bitcoin will be forthcoming soon.

k-of-n is out of scope for the paper. MuSig is only n-of-n, but with additional Bitcoin specific things, can be done in a k-of-n fashion.