The linked article [link to article in the Sunday Herald now broken, grr] indicates the poor level of debate on this topic in the popular press. Archbishop Conti (of Glasgow) is trying to make a somewhat complicated point. I'd like to summarize some of the arguments doing the rounds.
First, the natural law prohibition (according to the Church) is against contracepted sex. In terms of intentions it covers acts with (as Finnis calls it) a double intention: both the intention to have sex and the intention, if one has sex, that this act of sex should be prevented from resulting in pregnancy.
This gives rise to the following arguments.
Non-contraceptive uses of the Pill: for controlling menstruation etc.. The permissibility of this is uncontroversial. The permissibility of marital sex in this context is not affected - there is no contraceptive intention.
'Nuns in the Congo': can take the pill when fearing rape. There is no intention to engage in a sexual act. This sounds odd but is widely accepted in the Church. It depends of course on the strictly contraceptive effect of the pill (as opposed to the abortifacient effect).
Morning-after Pill for rape victims: an extension of the last argument. Finnis has defended this, again on the basis of the contraceptive effect of the pill. Its permissibility depends on the intention, and an informed person will make use of this morning after pill after rape with a contraceptive intention (as opposed to an abortifacient intention) only in certain conditions, depending on the time of ovulation and the time since the rape. There is a good precedent for this in the classical discussions (see Henry Davis Moral and Pastoral Theology, Vol II, p171). But the conditions attached to it undermine its practicability.
Non-contraceptive use of condoms: to protect against infection. This seems to be Conti's point. The parallel with the non-contraceptive use of the Pill is inescapable. The fact is that all sorts of things can render a person temporarily or permanently infertile, notably pregnancy, lactation, disease and certain kinds of medical treatment, and even if these things have been entered into voluntarily they do not affect the permissibility of marital sex, assuming they have not been done with a contraceptive intention.
The lesser of two evils argument: if people are going to fornicate, commit adultery, or resort to prostitution, it is better they do so with condoms, because of the risk of infection, and the injustice done to the children who are conceived. This seems to have been Cardinal Martini's argument.
Some uses of this form of argument are uncontroversial: if a pregnant women is hesitating between an abortion and giving her child up for adoption, the agencies of the Church will encourage her to do the latter - despite the fact that doing so is a serious sin. (The child has a right to be brought up by his natural parents.) It is far preferable, of course, to abortion. The efforts of Catholic charities to discourage abortion actually encourage the abandonment of children, but that (non-intended) result is less bad than the alternative, which is a larger number of abortions.
Other uses are controversial: needle-exchange programmes for drug-addicts; efforts to make prostitution less dangerous and unpleasant; free clean drugs for addicts; etc.. It is always a matter of balancing the seriousness of non-intended harms, and has be considered on a case-by-case basis.
Countervailing arguments draw attention to the non-intended harms of the promotion of condoms for infected couples.
The material scandal argument. It would cause scandal for the Church to allow the use of condoms in a sexual context (blowing them into balloons is clearly ok), because for the simple the natural law prohibition is against using contraceptives, simpliciter. 'The Church is against condoms.' This sounds simple minded, but highly paid media executives are very simple minded indeed.
The infrastructure argument. Distributing condoms and educating people in their use will increase the contraceptive use of condoms to a disproportionate extent. Collaborating in the policies of agencies and governments who promote contraception as a thing good in itself, will be even worse.
The Russion roulette argument. Condoms are far from fail-safe in preventing pregnancy, and the same will be true of preventing infection. (On the one hand the virus has to enter the bloodstream, but on the other an infected person is infectious all the time.) For a married person who knows he is HIV positive to have sex with his uninfected partner with a condom, would be for him to play Russian roulette with his partner's life.
On reflection, whatever one may think of the principles appealed to in favour of a permissive attitude to condoms, and whatever balance of harms derives from the earlier arguments, the Russian rouletter argument seems to be decisive. How could it be right to subject a spouse this this appalling risk, and to do so without necessity, and repeatedly?
Post Script: since writing the above I have encountered a different approach: that 'condomistic' intercourse is itself wrong, even within marriage, even with no contraceptive intention, because the condom barrier prevents it from being an act of mutual self-giving. This is an attractive argument, and is set out by William May here.