I've been on both sides but left the Intel business as rapidly as I could for two reasons. I knew I was not and would not be a good analyst (too impatient at the time among other things) and the sensing that I had entered an extremely cautious community that was more concerned with its reputation than its credibility; that seemed backwards to me.

As an operator and small unit leader and later a 3 type at several levels, my net experience over many years was that about half the Intel provided was adequate, accurate and timely; the other half tended most often to suffer from one of two failures. It was either over reliant on technical means and thus (in my day) invariably out of date or OBE. The second problem was that it often had been hedged to be safe so there was likely to be no error or failing on the part of the provider -- that also rendered it mostly irrelevant all too often. Additionally, the opposition had a vote and they'd do the unexpected or just strange things fairly often. Even the weather could intrude...

The bulk of the good and accurate stuff was on the more important or more dangerous operations. Unfortunately, that meant that the routine stuff was usually not that good. If you're an Infantryman or SF Soldier, it's the routine stuff that gets dangerous...

Seemed to me that unit S2s were often mediocre by default and that's not smart, when the 2 is a MI type, he should be learning what combat units do an embed the culture; many seemed to hate their tour in units and believed they could and would learn little (that also sends a bad message to the folks in the combat unit). The MI Dets and Bns had some really sharp folks but they were (logically) loyal to their Chain and sometimes gave short shrift to the supported units.

Too many Intel people tend to discount Humint and reports from units; they used to get an 'F6' more often than they should have. When the LRS Companies were placed in the MI Bns, the culture shock was tremendous (on both sides...) and the sad thing was that the MI people gave their 'own' LRS troops no more credence than they gave a distant rifle company.

That's the bad news from my perspective; the good news is that most of the time the system coped reasonably well and it worked more often than not. Personalities make a big difference. Some combat unit Commander believe in Intel and respect the providers, others do not. Most reach an accommodation. The type of war or operation makes a big difference. So called COIN like ops place a heavy Intel burden of a micro type and finite detail while MCO emphasizes the macro stuff and speed (simplistic but you get the idea), there are significant difference, operations type dependent.

The cultural aspect is difficult and my comments are from some time ago. My belief is that most combat troops realize the value of intel, are willing to accept they can sometimes not know how things are or were obtained and are willing to accept classification as necessary evil. However, they do tend to look at most -- not all -- Intel folks as REMFs who rarely provide timely and adequately detailed support and they tend to believe that much is vastly overclassified. The tendency is to replicate the old Poster, "14,000 Attaboys are wiped out by one Aw $*#t."

There is no evidence I've seen of conflicts in collection other than the mentioned tendency to discount Troop unit reports and Humint in general. There was some on analysis, generally relating to timeliness and to excessive caution and dissemination was very much product dependent with excessive classification delaying the product being a frequent complaint. That seems to be improved today. I hope it is.

As for:
if we are in many ways making our job more difficult through bias, perceptions, and expectations of the role, nature, usefulness, and methods of intelligence work.
I think that varies considerably from unit to unit, personality to personality and issue to issue as well as from operation to operation -- or war to war.

Some units want things they cannot have, most make reasonable requests and have logical expectations, some do not even know what questions to ask. Some MI elements do the best they can to provide outstanding support, some will sluff -- most do what they can with the resources given. In my experience, effective communication generally removed obstacles and made the system work as best it could at that place and time. The really bad things I saw almost invariably revolved around a "I cannot discuss that you don't have a need to know" sort of attitude in many variations. Going the other way -- as it did half the time -- the "I need this and I need it now; who the h3#* are you working for? The enemy?" thing was equally bad. Variations on both those themes...

Errors on both sides will occur but they can be ameliorated with decent effort and communication. Hopefully someone more current than me can let you (and I) know more...

Since I'm from Kentucky -- and it's a cold yucky February -- I'll take no umbrage at your location tag. That and the fact that I'm in Florida and it was 60 yesterday...