Personally, I see very little utility in using lower tier boosts. This is particularly the case with the ghodium boosts. Consider the Ghodium Oxide family which reduce damage taken by TOUGH parts. This is the most extreme example, but a similar cost/benefit argument applies to every other boost.
For T1 you need to mine or trade 5 minerals and carry out 4 reactions.
For T2 you require 2 more minerals and carry out 2 more reactions. For 40-50% more materials/work, you go from an effective 43hp bonus on a tough part to a 100hp bonus on a tough part. This is a 130% improvement for 40-50% more minerals. Given the choice between T1 and T2 tough part boosts, I would prefer to use half as many boosted tough parts and have a larger effect while spending less minerals and carrying out less reactions than I would on 2 separate T1 tough parts. T2 >> T1.
For T3 you require 1 more mineral and carry out 1 more reaction. For 14-16% more materials/work, you go from an effective 100hp bonus on a tough part to a 233hp bonus. This is again more than doubling in effectiveness compared to T2, but with an even lower incremental cost than it was to go from T1 to T2. Therefore if I have a choice between a 2 T2 tough parts, or 1 T3 tough part, I will always choose the T3 tough part as I will spend less minerals, carry out less reactions, and have a stronger effect. T3 >> T2.
But X, H and O are usually more expensive on the market and possibly less prevalent on the map
To an extent this is balanced by the relative prices (and possibly abundance) on the map. Only, however, to the extent that minerals are acquired by trade. Where minerals are mined by a player it takes exactly the same effort to acquire an X as it does to acquire an L or an O. So a player who expands his empire in such a way as to control rooms with minerals in proportion to his usage is not exposed to this form of balancing by cost. So I would argue that it's effect on the game is quite low.
Then how should these be balanced?
At the moment, I perceive no trade-off in always opting for T3 boosts other than more complex logistics chains.
IMO the effects of the increased tiers of boosts should suffer from diminishing returns, or the mineral cost and lab processing effort should increase at a faster rate than the benefits of the extra tier. Even if the mineral cost and lab processing effort scaled exactly in line with the benefits, the higher tiers would still be preferred over the lower tiers, because you can cram more boosted parts into one creep. It's usually more challenging for opponents to deal with 1 super-creep than 2 less powerful creeps.
With dimininishing returns on the cost-benefit ratio of higher tier boosts, players would have a more interesting decision to make. (Or more accurately, they have a decision to make for the first time).
Rebalance TOUGH boost on effect strength?
Recall that a T1 tough gives +43hp, a T2 tough gives +100hp and a T3 tough gives +233hp, we would need to ensure that the benefit of the higher tiers is less than that of the T1 tier. So moving from T1 to T2 should give no more than +43hp and probably less (due to the benefits of compactness, and less additional minerals and processing). I suggest +75hp in total for a T2 boost (a damage reduction percentage of 43%). Likewise a T3 tough part should be worse than three T1 tough parts which means that no more than another 43hp on top of the T2 boost. Because the incremental cost of the T3 is so much less than that of the T2, the benefit should be even lower. I propose merely another 25hp for a total of +100hp, or 50% damage reduction.
To recap: T1 remains as it is a 43hp boost / 30% damage reduction T2 moves to a 75hp boost / 43% damage reduction T3 moves to a 100hp boost / 50% damage reduction
Rebalance TOUGH boost on mineral cost?
If we instead leave the strength of the effect the same, we need to increase the amount of minerals used. I propose that we do this by doubling the boost from the previous tier (which would require either using 3 source labs in a job or creating new compounds).
T1 boost remains as it is and costs 5 minerals and 4 reactions. T2 boost would become OH + 2 * GO. This would cost 7 more minerals and 6 more reactions. T3 boost would become either X + 2 * GHO2 (8 more minerals and 7 more reactions) or X + GHO2 + GO (6 more minerals and 5 more reactions).
Are TOUGH boosts an outlier?
Yes they are. Almost every other boost has a benefit that accrues additively with each tier (e.g. ranged attack gives you +100% at T1, +200% at T2 and +300% at T3). While TOUGH boosts increase effective HP roughly multiplicatively each tier. A T1 tough part is equivalent to 143HP (1.43x an unboosted part). A T2 tough part is equivalent to 200HP (~1.43x the benefit of the T1 boost). A T3 tough part is equivalent to 333HP - 1.67x (!) the benefit of a T2 tough part. This makes the T3 tough boost stronger than any other boost in the game.
Even though the tough boosts are an outlier because they improve so strongly with higher tiers, almost all the other boosts suffer from a low incremental cost as you go up a tier and gain a large benefit.
There are two other outlier boosts though - Lemergium Hydride and Ghodium Hydride. These are outliers in the opposite way because they have diminishing returns - their benefit goes 50%, 80%, 100% as you go through the tiers which is a smaller improvement each time. These are used for repairing ramparts and for upgrading controllers, so I suspect the line of thinking that led to these diminishing returns is that they form part of the 2 most common end-game energy sinks.
Is locking boosts to room tiers the answer?
It would no doubt lead to more usage of lower tier boosts, however it would do this by reducing the options of players and forcing them to use lower tier boosts, rather than giving players trade-offs about which boosts to use. Because of this I would prefer to rebalance the cost or benefit of all of the boosts so that the best answer isn't (almost) always T3. Choosing to use T3 boosts should be a much tougher (ha!) decision than it is at present.