I don't think that when the mission planners and engineers sat down to decide what goes on the rover, things were ruled out because they cost too much money
They absolutely do.
Let's clear up some misconceptions.
- NASA has a budget.
- We can examine samples on Earth far better than any robot could.
- After some spectacular failures, the US has gotten very good at landing on Mars.
Vision and Voyages for Planetary Science in the Decade 2013-2022
Specifically, are there any primary-source documents or write-ups from NASA which detail their decision-making process in the conception of the Perseverance Mission?
Every 10 years the National Research Council (now National Academies of Sciences, Engineering, and Medicine) asks planetary scientists what they would like NASA to research and then recommends missions for the next decade. Vision and Voyages for Planetary Science in the Decade 2013-2022 decided a Mars sample return was the highest priority.
That document will provide the details you're looking for. Though it is just a recommendation it carries great weight with mission planners. The whole document appears to be available online for free.
Why Sample Return?
Scott Hubbard, NASA’s first Mars program director, sums it up in Mars 2020: Its Origins, Science and Technology.
Bringing samples back to Earth is critical for three reasons that have stood the test of time: utilizing instruments that cannot be shrunk to spacecraft size; engaging hundreds of scientists across dozens of laboratories; and most importantly, being able to follow the pathways of discovery as new experiments are conducted. As capable as Curiosity is, the instrument suite is fixed.
When designing a probe the mission planners have to make an educated guess about what instruments will be most useful. If they discover something they didn't expect, they must wait until the next mission to send fresh instruments. A sample return allows anyone to propose any experiment using any instrument.
Vision and Voyages Chapter 6 - Mars: Evolution of an Earth-Like World has a section on the Importance of Mars Sample Return. Here's some excerpts.
The analysis of carefully selected and well-documented samples from a well-characterized site will provide the highest science return on investment for understanding Mars in the context of solar system evolution and for addressing the question of whether Mars has ever been an abode of life.
Two approaches to the study of martian materials exist—that using in situ measurements and that employing returned samples. The return of samples allows for the analysis of elemental, mineralogic, petrologic, isotopic, and textural information using state-of-the-art instrumentation in multiple laboratories. In addition, it allows for the application of different analytical approaches using technologies that advance over a decade or more and, most importantly, the opportunity to conduct follow-up experiments that are essential in order to validate and corroborate the results. On an in situ mission, only an extremely limited set of experiments can be performed because of the difficulty of miniaturizing state-of-the-art analytical tools within the limited payload capacity of a lander or rover. In addition, these discrete experiments must be selected years in advance of the mission’s launch. Finally, calibrating and validating the results of sophisticated experiments can be challenging in a laboratory and will be significantly more difficult when done remotely.
They go on to give examples of inconclusive results from in-situ analysis and more detail about the return.
Mission Selection and Budgets
NASA has a limited budget decided by the US Congress. Missions must compete to provide the most science for the buck. Vision and Voyages made it their first selection criteria. From their Executive Summary...
To assemble this program, the committee used four criteria for selecting and prioritizing missions. The first and most important was science return per dollar. Science return was judged with respect to the key science questions identified by the planetary science community; costs were estimated via a careful and conservative procedure that is described in detail in the body of this report...
Mars Sample return is their top priority.
The highest-priority flagship mission for the decade 2013-2022 is the Mars Astrobiology Explorer-Cacher (MAX-C), which will begin a three-mission NASA-ESA Mars Sample Return campaign extending into the decade beyond 2022.
A major accomplishment of the program recommended by the Committee on the Planetary Science Decadal Survey will be taking the first critical steps toward returning carefully selected samples from the surface of Mars.
But cost was a major concern.
At an estimated cost of \$3.5 billion as currently designed, however, MAX-C would take up a disproportionate share of NASA’s planetary budget. This high cost results in large part from the goal to deliver two large and capable rovers—a NASA sample-caching rover and the ESA’s ExoMars rover—using a single entry, descent, and landing (EDL) system derived from the Mars Science Laboratory (MSL) EDL system. Accommodation of two such large rovers would require major redesign of the MSL EDL system, with substantial associated cost growth.
They recommended it be cut down to \$2.5 billion to have a chance of obtaining the money from Congress.
The committee recommends that NASA fly MAX-C in the decade 2013-2022, but only if the mission can be conducted for a cost to NASA of no more than approximately \$2.5 billion FY2015... It is likely that a significant reduction in mission scope will be needed to keep the cost of MAX-C below \$2.5 billion.
It turns out even that was optimistic. MAX-C was cancelled due to 2011 budget cuts putting a cap of \$1 billion on any flagship project far below what the NRC recommended.
Out of the ashes came Mars 2020, originally announced at the end of 2012 with a budget of \$1.5 billion. It would cut costs by building on Curiosity, sometimes literally by using spare parts. The eventual cost is \$2.8 billion over 10 years with \$2.2 billion of that for the rover.
In December 2013 NASA put out an Announcement of Opportunity for Mars 2020 "to solicit proposals for Mars 2020 surface-science investigations and exploration technology investigation". I'll leave that to be read for details.
NASA received 58 proposals (that's a lot) and selected 7 at an estimated cost of $130 million to develop and build.
NASA had to select instruments which would provide the most science for their goals using the limited money, mass, power, and volume available. Mass is the greatest challenge of spaceflight. Delivering 1000 kg of rover requires 531,000 kg of rocket and fuel and 900,000 kg of trust. Because of the Tyranny of the Rocket Equation each kilogram of payload mass can add dozens or hundreds of kilograms of rocket, fuel, and landing equipment which all cost money.
More instruments mean more payload volume. Space is vast, but space inside a payload fairing is limited to roughly 5m in diameter and about 15m in height (the Atlas V 541 used to launch Perseverance has a very tall payload faring, but includes the 13m upper stage). There's only so much which can be crammed in, and some simply cannot fit at all.
Finally, more instruments means more power and power is very limited on Mars. Perseverance uses a very reliable Multi-Mission Radioisotope Thermoelectric Generator (MMRTG) providing 110 Watts, barely enough to run a desktop PC, supplemented with batteries, and it only gets weaker. Spacecraft have flown with more powerful RTGs, Cassini-Huygens carried 3 RTGs delivering over 800 watts at launch, but that adds cost and mass.