Far from being a "a detailed description of how to do it", my post is a very superficial description that leaves out all the details of actually doing this. Trust me, as someone who did this kind of thing for a living for decades, there are a lot of details to actual implementation of any study, and as they say in the military, "No plan survives initial contact with the enemy".
Research needs to be reviewed and approved by human subjects protection committees (Institutional Review Boards) to make sure that studies do not harm the people that they are studying.
As to folks who don't want to be counted, this is always true with any hidden population- this is what makes them hidden. There are lots of statistical methods for making adjustments to account for this.
On a national scale, this is a multi-million dollar/year project- no one is going to pay for it any time soon. A very rough cost estimate, based on doing similar work for many years, is $500,000/year in each of 25 cities, for a total annual cost of $12.5 million, and then another couple million to analyze the data generated by the 25 cities, so maybe $15 million total. Read my last paragraph again. Doing "recapture" interviews to get any kind of national sample would be very very challenging. If i were doing this, i would definitely be hiring folks with experience "riding for about 5 years solid" as my staff, along with a bunch of statisticians/epidemiologists.
also, Beegod Santana, if you would like to read the actual Hobo Code generations of nomads have lived by, google "Hobo Code 1889".
How would a hobo find work if he/she/they was invisible? Being "never seen" sounds more like some criminal tramp shit than being a hobo. No shaming intended- i was a teenage parasite on society, but if you aren't looking for work when you travel, you ain't no hobo.