I'll go out on a limb...
and suggest the hardware WAS capable of doing all the things they wanted to, but poor management, improper development techniques and inadequate testing were the root cause.
I know they have made at least some improvements via automated test methods, what I don't know is the caliber of those currently coding the unit as they lost 2 very good programmers a few years back along with a spirited program manager with real vision and drive to have it reach it's full potential.
Why management at the time would see fit to derail a program is beyond me.
If I'm wrong, then everybody wins. (Everybody should hope I'm wrong)
If I'm right, then nothing has really changed, and from lesson learned from M$, throwing more CPU and memory can only cover bad coding and design for so long.
I believe the fundamental architecture is flawed.
Unless the 6XX is truly a rewrite from the ground up.
Along with some fundamental H/W changes as well.
But given the investment made for legacy code to GIST (Blue Wiener Bar Guide) and licensee fees for OpenTV, I can't see them walking away from all that.
The problem is Shaw continues to make money in spite of it self, but it's stock would do a lot better if they really got behind a program as they have done in other areas and become market leaders in all business divisions.
The re branding to Shaw Direct is a slap in the face to what was once a good and respected name in the marketplace, but the unit appears to still be viewed as the ugly daughter of the family.