::"Should I fully enclose the back of the rack by building it out of a solid sheet of something, or leave it more open by installing narrow diagonal supports that provide solid structure and allow maximum front to back air flow?"
With an open back and heat tape behind the boxes there will be a more significant loss of heating efficiency rather than relative decrease in airflow. But it's not because of heat "rising" necessarily. The heat produced by heat tape is mostly radiant heat which moves similar to light rays. It does not really "rise" like convective heat. For reference, heat travles in three major ways: radiant, convective, and conductive.
But "belly heat" still tends to be more efficient as there is more control of the direction of the radiant heat rays. And there is some convective heat produced and these will rise.
Whether the back is open or not the box will be the limiting factor in air flow. That is unless you drill A LOT of holes that are both high and low on the box. Many of the gases in herp cages are heavier than air and will settle in the bottom of the box. Not a big deal provide they are openended several times per week.
::"Of course how much airflow is there under ground or deep in a rock pile where these guys usually live?"
Quite a bit more as the heavier gases are more likely to escape.
::"I could hedge my bets and use peg board on the back which allows some airflow."
I think a better idea is to have a lot of holes drilled on the front sides of the boxes. This will put most of your ventilation near the front of the rack where you will have the most airflow - even if you have an open back.
On the back attach a 1" piece of that foil-covered, styrene based insulation. The foil blocks most of the radiant heat rays produced by the tape, the styrene slows down the loss of convective heat.
With this design there *might* actually be some advantages to installing the heat tape behind the boxes rather than underneath. To describe the differences or possible advantages is fairly complex and would depend on several factors.