Following my introduction with this post, the next step would be to clear out the romance and attack spell research as a design problem. Specifically, what are we designing, and how do we get there?
To explain that, I have to repeat my position on how spells work. A "spell" is the assembling and ordering of natural forces, however misunederstood by a non-magic using society, in a specific way so as to call a predictable effect. This assembling is accomplished, in my game, by producing sounds and moving the body, while concentrating the mind, through a period of time that a given spell requires. More powerful spells are more complex, so that the assembling of the natural forces involved takes longer. Some spells may require hours to assemble this power.
Assembling these forces has led to different strategies by different classes. Clerics pray to appease higher powers than themselves, so that they may rely upon these higher powers to intervene in the magic's assembly when the moment comes. Druids seek innate energies within the world's physical space, which they understand and can use at the right time to assemble their spells. Bards do something similar, but they weave spells with music. Finally, illusionists and mages rely entirely upon their own minds, assembling spells with considerable mental acumen.
If we are researching a cleric spell as opposed to a magic or illusionary spell, the methodology is entirely different. Therefore, a spell-researching system must be flexible and view the problem from more than the static angle of acquiring materials and laboratories. Why would a cleric need a laboratory to speak more nearly with the cosmos? Why would a bard? Or a druid?
Moreover, we need to consider the point of reference between the spellcaster and the availability of the spell. A mage or illusionist studies a spellbook, which orders vast amounts of information into the character's thoughts, which the character spends time organizing, like a memory expert creating a thought-cathedral in their mind. The casting of the spell shatters this order, which is tenuous and is, in large part, subconciously maintained. Therefore, it must be reordered again before the spell can be cast. The term, "memorizing," is merely a placeholder. It is a convenient word to describe something for which we have no word.
The cleric spends time effectively pleading with the cosmos, asking, "Please let me cast this spell again today, I cannot do it without your help." The druid mediates and invests self with the ever changing environment, as the various energies that are everywhere shift daily; once those energies are found, they are tapped, and the druid stores spell energy in his or her body, which can then be released; but it has to be found again the next day, after resting. The bard practices, tunes the instrument, spends as much as an hour finding the perfect tone and resonance, which takes time and effort to do ~ then affixes a perfect memory of that tone so that it can be played later that same day. The next day, a change in the weather, the age of the bard, the stiffness in the bard's fingers, can all mean time spent needed to find that tone again.
More precisely, then, the mage and illusionist are looking to create symbols in a spellbook that can be, in turn, studied and used to order the new spell in the mind. The cleric seeks to appease the god into giving a new spell that has never been granted on Earth. The druid requires new knowledge of the environment. The bard requires a new song.
Before the symbols can be written, they must accurately describe the manifestation that is to be created. The cleric must be able to explain precisely to the god what is needed, and convince the god to perhaps turn to other gods in order to gain the power, that can then be sent on to the cleric. The specific concordant element of the Earth itself, and perhaps the universe, must be identified and found by the druid before it can be tapped into. The song must be heard, perhaps in the bard's imagination, perhaps in actual fact, before it can be written and repeated. These are different journeys, but they amount to the same thing: discovery.
The path is an adventure. And like an adventure, there are many paths that might seem like the right one, but the DM already knows, before the players start out, what the right path is. The DM may provide ideas, or clues, or proposed strategies, presented in books and out of the mouths of experts, but the DM knows which experts are lying and which are telling the truth.
Like moving through a dungeon, the players have multiple doors that may lead them to a spell. Some doors are false. Some are real. Some will lead to dangerous, but profitable outcomes; others will lead to simple, but fruitless results. As each path is tried, and discarded, the player comes closer to the goal.
Let us take an example. Suppose the spell "dancing lights" does not exist, and the player would like it to exist. Our first question is, what exactly are dancing lights? We have the spell description, but that only tells us the result. It does not explain what the lights are, or how they manifest, or what they are made of. Clearly, not fire.
The DM creates three pathways. That is a lot of work for some DMs, I know, but I am explaining what I would do with spell research, given my degree of experience, and personally I would find it quite easy in a few days to spontaneously create three pathways. In fact, I managed it since Friday:
One: Research elementism. Insert magical fire into natural fire, in a way that attempts to produce natural fire that does not need fuel to burn, using alchemy. By reducing the heat of the magical-natural fire, perhaps it can "burn" without producing heat. Experiment with other spells that allow telekinetic control on a very minimal basis (no weight, much easier than telekinesis).
Two: Using suggestion and ESP, experiment with the creation of thought-creation in the minds of test subjects, to see if the dancing lights can be impressed in their consciousness. Explore mass delusions, as well as spells such as massmorph and hallucinatory terrain. Perhaps either can be simplified through painstaking work to draw a 1st level spell out of a 4th level spell's design.
Three: Begin with the light spell. Build a prism-based construction that will split the light into separate pieces, perhaps employing elements of the mirror image spell. Using physics, split the lights in some manner that causes them to flicker, producing only the yellow, red and orange parts of the spectrum. Since the light spell can be positioned the four weaker dancing lights should be likewise able to be positioned, and then made to move in some fashion.
Each of these is designed to produce the effect first. Once the effect is produced, the mage and the illusionist can scribe the complex description of the effect into their spellbook, so they can master the effect when they need it. The bard can hear the lights, and produce the music that will create them again. Druids and clerics come up short. The spell is not available to their disciplines.
It would be a mistake to think that all three of the methods above will ultimately work. That would ruin the game. The player's frustration and sense of meaning in the exercise depends on one, and only one, means to the truth. This makes the truth valuable ~ and their decision-making, as they posit other spells, existing elements, related concepts, etc., into their efforts, entirely of their own making.
Spell research is making something out of pure imagination. It defies ordinary rules for that reason. But design follows specific, ordered pathways. Propose the idea. Brainstorm a means to get there. Experiment with each means, to see what results. Produce the result. Reproduce the result again. Write it down and make it standard.
The rest is the work of thinking, on both sides of the table.