In the Wealth & Poverty chapter of Covenant (p.67), it is stated that a magic item can reduce expenditures: "As a very simple rule of thumb, each magnitude of effect in such items can save one pound per year."
But a case could be made for the same item to provide a laboratory virtue with no increase to Upkeep, such as Magical Heating or Magical Lighting (p.119), which may mean greater cost savings.
So the question is, can the same item apply to both? Can a magic item even apply to multiple labs?
Let's take an example. In a recent thread, a spell was proposed to create olive oil for lamps. Let's call this a level 20 item (Effect level 10, +10 unlimited uses). Turned into a magic item, this could provide essentially unlimited fuel for lighting.
As a covenant cost-saving measure, according to p.67, this would save up to 4 pounds per year (1 per magnitude) as a rule of thumb.
But a case could be made that the item supports the Magical Lighting for one or more labs. This does not reduce the Upkeep of the lab, per see, but prevents the increase of the Upkeep by 1. Just to remind you, raising the Upkeep of a lab increases its cost along the pyramidal scale. So if it rises from +2 to +3. Points for that lab go up by 30, which increases the lab expenditures by 3 pounds per year for each lab.
So should the item provide only 4 pounds of cost savings? Provide the Magical Lighting for one lab only? For multiple labs? All of the above?
We allow anything that makes sense Those given are just rules of thumb.
As for labs, we tend to be really strict. In your case, we would allow the free olive oil to cut the Upkeep costs of every laboratory where it's used to provide excessive/superior lighting/heating. But it would not provide the virtue Magical Lighting or Magical Heating automatically. You still have to spend a season (and money) to outfit the lab, buying the lamps or the stoves, positioning them (and they take space!), taking all measures to avoid accidental fires etc. Incidentally, 1use/day, for a final effect level of 10 rather than 20, is more than sufficient: 4000 liters oil/day can light and heat a LOT of rooms.
As for expenses, we just figure it all out on a case-by-case basis. All that olive oil would provide little help to a covenant located on the coast of Sicily, where even in winter you hardly need any heating at all. It would help somewhat with the illumination and with the cooking, but not much. A covenant in on some wind-blasted alpine mountaintop would be a different story; we would allow the oil to provide the maximum benefit of a craftsman in the consumables category (a 20% reduction of the corresponding expenses). It helps a lot that the covenant's expenses are divided by type (wages, consumables etc.); we just eyeball which types of expenditures are targeted by some cost-saving magic, and by how much, and that's it.
Incidentally, we allow cost-savings even from simple spells, not just magic items or Rituals. This means that our troupe is constantly trying to find ingenuous uses of known spells, which we rather enjoy. The first catch is that, if they entail botch dice (e.g. fatiguing spontaneous magic), then we automatically slap 1 Warping point onto the casting magus for every 100 castings (about 1 Warping point/season for a spell cast daily, 1 point/7 years for a spell cast "Moonly"). This means it's mostly formulaics that the magi had to invent, or non-fatiguing spontaneous magic of first, and in some rare cases second, magnitude. The second catch is that if the spell must be cast periodically, and the magus is away (at Tribunal, in Twilight etc.), then the Covenant incurs all the consequences...
While it may not be a terribly helpful response, it's things that this that are why we ignore the whole financial book keeping aspect of the Covenants. We just comfortably hand wave it, using the guidelines from the core rules. Hasn't detracted from our enjoyment of the game one bit.
Just treat the "guidelines" in Covenants as guidelines. If a particular item/spell can do multiple things, or saves more money than is typical, or whatever (given the context of your covenant/saga), then it is just a particularly good item. The covenant council will give a hearty three cheers for the clever magus responsible, and may think about selling copies to other covenants.
What we do is handwave the covenant finances, but impose an upper limit on the amount of upkeep a lab is allowed before it is deemed 'too pricy' for the covenant.
As an option (we've not exercised it because no-one has expensive labs), you can always have anyone who exceeds the limit acquire the 'expensive lab' story flaw - meaning they will prompt stories on their lab being overpriced.
This lets us use the lab rules and have upkeep mean something without having to worry about the finance rules.
What the maximum upkeep is depends on how wealthy the covenant is. +1 for an average covenant works fairly well. Increase/decrease based on wealth boons/hooks.
I like the old way of keeping money out of Ars Magica games. Obviously this shows my bias towards wizardly stories and not companion/grog ones. I'd rather wing it and come up with stories as to how your characters are actually going to get that fancy glassware to the wilds of Scotland or how they're going to keep up their supplies of those rare herbs that only grow in Syria.