Good info! I think that limit is applicable to the commercial/consumer goods we're discussing. I'm rather certain that semis have a higher rating for industrial and military applications, but I don't pretend to know how that is attained.
I wish I had a recorder to measure the temps. of interest over time and under different conditions. Regardless, I think we have enough evidence that the current design operates these BC chips near their limits and that can't be good for long term reliability. I have to wonder how many "glitch" problems reported with the 622 are caused by excessive heat in these chips. There may even be some hard failures that could have been avoided.
I still need to follow through with the rest of my "project" as outlined in my aforementioned post. In the interim, those chips are running a good bit cooler with the cover off and 24/7 forced cooling and I have to think there is no downside to that other than the minimal upfront cost of those fans and the marginal extra power consumption. The HDD might be a bit better off as well...