This might be a good point to re-cite this article from Blue Jeans -- one of the better regarded makers of cables -- as to why HDMI cabling is so tough to get right, and in particular why adapters (including wall plates) add to the problems as the bandwidth of the signal you are passing goes up:http://www.bluejeanscable.com/articl...-with-hdmi.htm
I think what some people may be missing is that this stuff is thorny enough that it may even come down to batch differences in the same model of cable/adapter from the same manufacturer.
The people responsible for the HDMI spec are well aware of the potential for problems of course. But they know that properly engineered and tested cables and adapters WILL WORK. And that's why they pushed through the much stricter design and testing standards for cables that want to use the HDMI V1.3 label.
They are also banking on a new generation of HDMI V1.3 (and higher) chip sets which will include much improved signal "equalization" processing that enables the product to work reliably even in the face of signal degradation by the cable (and adapters). The problem is, there is no way for the customer to tell that a given HDMI V1.3 product includes such chips. It is not part of HDMI V1.3a or even the latest HDMI V1.3b labeling for products.
What is clear is that older HDMI V1.2a or lower products can not have such improvements, and that DVI products are worse at this than even HDMI V1.0 products. Which means if EITHER END of the cable is attached to such a product (or even an HDMI V1.3 product without the new, magic chips) you need to be more concerned about total cable length from product to product and the quality of the cable(s)/adapter(s) in the path. This is particularly tough on folks who have a DVI product at either end and want a cable run of more than, say, a mere 10 feet. And folks using 1080p/60 video are MUCH more likely to run into cabling issues than folks using 1080i or 720p video. (The even higher resolutions defined in the HDMI V1.3 spec for future products some years down the road will only make this worse.)
Anecdotal evidence that someone has managed to make a really long cable run work even through adapters just means that sometimes you can luck out. Another person using the very same electronics, cables, and adapters may not be so lucky simply due to manufacturing variances and the quality of the plug/socket connections.
Note that this is just between any two products since HDMI and DVI products, except for the simplest/cheapest passive switches, invariably regenerate an input signal to create a new output signal -- thus starting the process over again.
This also means that if you have long-run cabling problems, the solution may be to add an HDMI repeater of some sort (often a switch) into the cable path. But that can only switch the problem over to the other bugaboo of HDMI which is that each product added to the TOTAL signal path from source to display complicates the job of the source in verifying HDCP (copy protection) from end to end.
The fiber optic (and new coax) in-wall HDMI cabling solutions solve this by essentially putting a piece of electronics at either end of the in-wall cabling and then using fiber (or coax) in between them to eliminate signal degradation issues over the long run through the wall. The box at each end has the effect of repeating the signal. Again, this should eliminate signal degradation problems but MAY introduce HDCP problems.