Estimating Subcutaneous Fluid Dosages by Body Weights
The primary philosophy behind sub-q's is maintaining hydration. Since the fluid is not going directly into the bloodstream, it is not really exerting a "flushing" pressure on the kidneys the way IV fluids do. When we speak of "flushing" relative to sub-q's, I believe it is that we are bringing the kidneys more toward their normal function than when dehydrated. In theory, excessive administration of sub-q's are not absorbed into the circulatory system, they just pool and are absorbed later when needed. In practice, they do go into the lymphatics and into the interstitial space, which has its own mechanism for other metabolic tasks. In other words, it is possible to overhydrate, though not an immediate danger in most cases (barring heart disease).

To answer the real question: how much fluid to give? (I'll give the whole story, the short answer is at the end.)
The goal in determining dosage is to balance total daily fluid intake with total fluid output. According to the vet manual I use, intake includes oral (food and drink) and metabolic (generated in the body by the metabolism of carbohydrates and fat). Output includes urine, sweat, feces, insensible, and water incorporated into new protoplasm. As a guideline, the book lists the normal daily water turnover value (intake=output) as being about 42ml of water per lb. of body weight. This is broken down in the NORMAL cat as follows:
Oral (Food and drink) 36 ml/lb
Metabolic 6 ml/lb
Urine 20 ml/lb
All other 22 ml/lb
Obviously, to follow all factors accurately would be impractical, for instance weighing sweat and insensible losses through the breath. Even measuring urinary output accurately is not generally possible, and weighing food, etc. would probably not be worth the effort. With Coco, we developed a simplified system that seemed to work.

First, assume that the sweat and insensible are probably insignificant in the order of things, that metabolic input probably approximates loss to new protoplasm, and that moisture in food probably approximates moisture in feces. This leaves a more manageable set of variables to track: water drunk versus urine. Sub-q's should supply the net difference.

Rather than measuring urine (hard to get all of it) and water (how much lost to evaporation?), it is more practical to use a series of body weights. If it is assumed that Tom's body mass is about the same in the morning as in the evening, and the same as the day before, then it follows that any fluctuations in weight are due either to eating, drinking, urination, or defecation.

What we did with Coco was to establish a nominal base weight, and measure the sub-q's to keep her at that base weight when fully hydrated. (We used a postal scale that reads up to 10 lbs, with a tray for her to sit on that weighed 1 lb 3 oz, which we subtracted from each weight. The scale is marked in ounces, we estimated and recorded to within 1/2 oz.) The starting base weight was determined relatively arbitrarily... we picked a time to weigh her about halfway between sub-q's, so that she was neither dehydrated nor overhydrated, and neither immediately before or after eating, urinating, or defecating.

We kept a log of weighings, with notes indicating whether they were taken immediately before or after a treatment, before or after urination, etc. We weighed Coco whenever she was due to urinate, and immediately after she did. (After a while, it became apparent that Coco would normally void about 1 1/2 oz. If she did so four times in a day, that represented 6 oz or about 150 ml.) We weighed her before and after each treatment. This confirmed the amount given, since as we all know the markings on the bags are impossible to read accurately. For simplicity, we used a conversion of 25ml per ounce of weight. This is a couple of grams off, but accurate enough to serve the purpose. Thus, if she weighed 5 1/2 or 6 oz more after her treatment than before, it confirmed that we had given her 150 ml. We also recorded the amount of each treatment. Having this information on paper helped us track her progress. If she weighed a certain amount one morning after urinating and defecating but before her treatment, and the next morning she weighed 2 ounces more, she probably had retained 50ml of fluid. We might take that as an indication to reduce that treatment by 50ml. The reverse was also true.

THE SHORT STORY: establish what you consider to be the normal weight AFTER TREATMENT to within 1/2 ounce if possible at a standard time of day, standard state of bladder (preferably empty), etc. Weigh before giving fluids, and give fluids at the rate of 25 ml per ounce deficit, in order to restore to "normal" weight.

Another Excerpt Follows:
(Question: Should a cat getting 150ml every two days switch to a daily dosage.)

The main reason I see for increasing frequency is to control dehydration. If the kitty doesn't seem overly dehydrated two days after a treatment, then that timing is probably ok. Depending on weight, amount of other fluid intake, etc., this may be the case. For a lot of people (but not all, there are no universal rules), it doesn't seem like it's worth it to stick the kitty for less than 100ml. If much more than 150 ml was needed at a time, I personally would consider spreading out the dose. Clinical dehydration evidences, I believe, at about 5% of body weight (correct me if I'm wrong). Rounding the measurements crudely, at 4 oz for 100 ml, this means that a 5 lb cat (80 oz) would be clinically dehydrated at 100 ml deficit. If the dosage was 200 ml every two days, after treatment, there would be a reserve of 100 ml. After 1 day, hydration would be "normal". After 2 days, when the next treatment was due, he would be clinically dehydrated. Switching to 100 ml daily would seem logical (starting one day after the last 200ml dose), in that it would maintain a level always at or above a "normal" state of hydration.