Haist
Last night I picked up Haist's Modern Photographic Processing from the Kansas University's Art Library, specifically Vol. II which covers reversal processing. What a book! I also picked up Wall & Jordan's Photographic Facts and formulas, also worth a read.
If you are remotely serious about RPing, read this chapter by Haist. It was an absolute eye-opener and is by far the most comprehensive discussion available, and it's fairly easy reading to boot.
I'll try to summarize the most pertinent stuff.
1ST DEVELOPMENT
The most surprising and enlightening thing learned is that Haist considers the 1st developer and the 1st rinse to be "the 2 critical processing steps of reversal processing, determing the photographic quality of the positives." The rinse is important because the active alkaline developer continues for a short time in the rinse bath, giving "desirable edge effects". Or, this can be eliminated/curbed by a stop bath, or probably by a more immediate and vigorous rinse.
That being said, nearly all control has been affected by the time the film enters the bleach bath.
The recommended 1st developer is Kodak D-67, which is identical to D-19 with the addition of 3mL/L of 51% liquid sodium thiocyanate. Wall/Jordan second this recommendation. Basically a hydroquinone developer with a smidgen of "metol of phenidone" (plus other important things that I forget at the moment) is ideal. Phillip T. Jones (?) thought this was too contrasy and proposed a 2-part developer that is available in Haist.
This silver-solvent is by all accounts very important in the 1st developer for results that are 'par excellence'. Haist and Wall/Jordan lay it out quite clearly, better than I can do. But basically, it has to do with physical development. Potassium thiocyanate is the preferred solvent at 0.2% according to Wall/Jordan and at 0.5% according to Haist. Above 5g/L and density goes down, below 5g/L and ISO drops and you get longer development times.
Contrary to popular formulas, hypo is not considered to be very good, resulting in lost shadow density.
Alternatively, "a more caustic developer like D-88 or D-8 permits shorter times, 10min @ 68°." This requires no silver solvent, and thus can be reused as the 2nd developer.
As for development time, here is a quote from Haist, "1st development, although not carried to completion is usually sufficiently long so that the contrast of the image is no longer increasing rapidly w/ time. Film emulsion speed, as determined by the least exposure to give clean highlights in the positive image, does not change appreciably w/ small variations in developing time. Thus there is some tolerance in the amount of 1st development caused by agitation, temperature and solution-exhaustion variations, but such deviations from normal must be minimized."
If shadow dMax is low, could be too much 1st developing, and vice versa (excessive density in positive = not enough 1st dev.)
Increased time in 1st dev. (pushing) can increase highlight film speed, but w/ a reduction in the dMax of the positive.
BLEACH
I paid greater attention to dichromate bleaches, so I recommend users interested in permanganate bleaches to consult the book. Interestingly enough, there is a 3rd option for a bleach, using ceric sulfate. Apparently it is expensive though, but interesting nonetheless.
A dichromate bleach of 0.5% P.D. & 0.5% sulfuric acid should have done its job in 1 minute, and the time should not exceed 3 minutes. IIRC, permanganate bleaches should not exceed 4 minutes.
For some films, a blacker tone & better tone gradation may be achieved w/ permanganate over dichromate.
Old dichromate bleaches can redeposit chromium and silver if the pH increases, which happens with exhaustion. This can result in stained highlights in the positive. It is minimized by using the lowest necessary 2nd light exposure, since these chromium/silver particles are light sensitive, but to a lesser degree than the silver-halide. If they aren't made developable, they will just simply fix out in the fixer. Also, reconstituting the bleach by the addition of sulfuric acid will do the trick.
Permanganate bleaches have no keeping abilities.
CLEAR
Wall & Jordan say to use a 9% sodium-sulfite clearing bath, and to avoid white light and times longer than 2 minutes because this bath tends to dissolve silver halide and thus you get a thin positive. Haist seconds the notion of dissolving silver halide, but suggests at 5% solution of sodium-sulfite or ideally a 1% solution, if time permits. Of course, sulfite only applies to dichromate bleaches. But the concentraions for bisulfite and metabisulfite are similar IIRC.
FOGGING
Sorry, I didn't record anything about fogging developers or the like. But, for light re-exposure, a recommended quantity in some old movie-film formulas was 800 fc/sec. Basically, re-expose sufficiently, but don't go bonkers with excessive light amounts.
2ND DEVELOPMENT
Any vigorous b&w negative developer, such as D-19, is suggested by Wall & Jordan. A small amount of anti-foggant (benzotriazole) will avoid stain and brighten highlights.
Haist says that good developers include D-72, D-8 and D-19. Bad developers include D-76, DK-50 and any warmtone paper developers.
FIX
Do it. Hardening fixers are not a bad idea, since the gelatin has been stressed in the bleach and with 2 developers.
--------------------------
Lastly;
dMax should preferably be at least 2.0, more like 2.2-2.4, with densities greater than 2.4 not contributing significantly to perceived density.
Haist says that thick emulsion layers w/ high silver coverage (like Tri-X) have been found difficult to RP. Also, fine grain, thin emulsions (Panatomic-X) may not have enough silver halide left for the 2nd developer to produce good density. Mind you, this is from 1979, so I'm not sure what films this might mean today.
The only wash that is critical is the 1st rinse, as described above. Other rinses will ensure longer solution life and will minimize inconsistencies but are less critical in terms of the final outcome.
...many different films, devs, both bleach solutions etc...