You know, I've never understood this line of argument. The way I see it, the client is either clever enough to see technology really is not important to the problem at hand or too dumb to realize it is. If the former, you'd be guilty of overengineering and missing your client's objective (by being too costly, too complicated, etc). If the latter, why would you care to explain and "do the right thing" (the client will not pay for things he doesn't know are important)?
There are many technical people who feel underappreciated by their clients or management peers. I think the only solution is to find out if the wizardry really is needed in the first place; if it's not, do the minimum (if you accept at all); if it is, then let the bozo fail and either compete (you know something they don't) or leave for a better client/customer/mgmt. Will pay better and you get to keep your sanity and pride.