I would say he is clearly wrong about the cost given the amount of states that do them, but might be correct on the data. Hunters overemphasize harvest. I have even complained a little that FWP overemphasizes harvest in their model. Using past harvest to adjust for future harvest is backwards. Ie, if you know you are in a drought, cut B tags THAT year. Don't wait until you count fawn ratios next Summer and see that a lot of does are without fawns. See my comment below on AHM for waterfowl.
HIP surveys are used to stratify the sampling method when FWS sends out actual harvest reports. You will notice they have wide brackets in harvest. There is no precision. They want to make sure they are sending harvest surveys to various people including those that hunt waterfowl occasionally as well as hard core waterfowlers. The entire HIP survey is to eliminate any sampling bias. The actual harvest surveys are used to estimate harvest. Here is the kicker - in waterfowl, the harvest has little if any impact on season setting (think of when you fill these out). It is all based on May pond counts and nesting counts. The know that water and cover on the nesting grounds = ducks. I am not even sure they do mid-summer brood counts anymore. Any future harvest is limited by adjustments in bag limits, but season lengths have generally been liberal since they went to AHM. Debating differences in Big game versus waterfowl are valid, but the general estimating process should fit. Also, they can do this because the model has 90 years of data to stick in the formula, flying and walking the same routes over 90 years and making sure any changes are both justified and accounted for in the model. It is expensive, but it works. Regardless, hunters still complain.