John & all,
My experience as a buyer (in bulk, I might add - hence the undisclosed name) of global industry research is that several factors are hugely important.
One factor to understand is where the researcher "plants their flag" when attempting market sizing measured in currency (e.g. dollars, Euros, etc.). Are they measring from the ultimate end customer's wallet (including equipment at integrator cost plus installer labor)? Are they measuring at the distributor's shipping dock (equipment plus distributor markup, with no dealer involved?) Or, are they measuring from the manufacturer's shipping dock (no distributor markup, no integrator, no labor?)
Another factor is HOW the researchers did the research... I can do a lot of my own legwork using public data, so for me it's a "make vs. buy" decision. Reputable research outfits are happy to disclose the precise methods used; others view their method as a trade secret. Ask: and if the company won't share the method to your satisfaction, run away. (If the company has to give away fake awards to get noticed, run faster!)
Also: having some understanding of what researchers actually do will help you make these decisions. I break it all down into two groups: Qual and Quant.
"Qual" (qualitative) work is when the researcher performs interviews about a topic and summarizes the findings - an example would be to interview ten integrators and ask them what sorts of jobs they expect to be pursuing in the coming year. Qual research is important, but computationally "squishy"... that being said, it's helpful when you are tackling squishy issues such as consumer behaviors or preferences. Qual research is also helpful in informing "quant" (quantitative) research: Qual tells you where to look, Quant helps you see.
"Quant" (quantitative) research is when the researcher gathers data of a numeric sort: how many integrators pursued Canadian government contracts in 2012, how much work (in currency) was granted in 2012, etc. Quant data also comes from bulk surveys - one well-known consumer electronics research firm sends quarterly Web surveys to something north of 30,000 broadband equipped households... which may not tell the researchers what the customers are dreaming about, but can very specifically tell what they are buying. Compare the researcher's methods to your need for information and make sure you're speaking the same language.
Quant data on global video hardware, by the way, is the toughest data to come by in this space, because of the OEM issues. (The truth may be Out There, but none of us have it yet ;-)
Hope this helps, and apologies again for the undislosed name.