Sorry I have to laugh, but those guys in Bulgaria remind me of a Deja Vu with another big company that seems to have found the same cost friendly country to place their support at. Those poor chaps acted often the same like you described and sometimes didn't want to pass calls to the next level which was also no big help.

So what Rob wrote in his answer, that you as customer should escalate etc. is in my eyes not a nice but maybe a common business behaviour these days with some big companies, to have the customer involved to ensure the quality of the vendor's support.
We heared this with the other big company too, but having to involve an escalation manager etc. gets tideous after some time as well and one asks himself, what is going wrong there, that I have to do so much effort to get some help or sometimes at least someone that even understands what my problem is.
If you put people there, that have not enough experience to offer good support, then this is a problem of the vendor and must not be a problem for the customer. It feels a bit like the concept of green banana software being used for support structures.

So why is the HMC so locked up...
Yes, in the long time as AIX admin I did not like it at all and absolutely agree with you, that people that are responsible for plenty mission critical servers with sensible applications/users, that already have the knowledge at hand to get along with the HMC, should be allowed to do so by default. They can still open up a support call if they get stuck.

Because if an admin has no clue and screws up one or many LPARs, he will usually be in more serious trouble than the one that screws up a HMC, which usually comes redundant with 2 of them, where not all important LPARs are always redundant.
And don't forget the VIOS - do something wrong there and you have a good chance that really lot's of LPARs get problems, so what.
In the end in a professional environment one will have a backup for the LPAR, VIOS as well as for the HMC.
And severe LPAR damage has most often a direct impact to users ie. our customers, even if it is "just" a cluster switch that takes some minutes but gets maybe 10k users disconnected and maybe some unpleasant attention by your boss/managers. Trouble with the HMC will usually go unnoticed by our users.

So the HMC is at least locked up for 2 reasons in my eyes:

a) The customer has strongly to rely on the support of the vendor. This is a dependency and some kind of "bonding" of the customer to the vendor. The vendor gets cash, the customer has a helping hand and feels good withit, simply they are just good friends and will most likely have more business in the future So far the possible theory.

b) It was said in the discussion, that the admins often have not enough skill/experience - true, but these guys have been in business way back in time and such will be in the future.
I have the impression, that in favour of cost efficient support structures, they have tried to make the HMC to be being easily maintained by their support, not because the customer side is so unexperienced.


cheers
zaxxon