Given the predicament, the research, trials and deployment of both modes of geoengineering seem to me appropriate under a range of limiting conditions.
The first, as Neven rightly remarks above, is the systemic change of society as an accompanying transformation - without which Geo-E is patently insufficient.
A second is of governance, with a UN scientific agency mandated for the supervision of proposals' research and trials, and potential accreditation, with any deployment being by the collective decision of the UN member states, with such decisions for the Albedo Restoration mode being valid only after a credibly stringent Emissions Control treaty is in operation.
A third is of the need for both modes' deployment alongside Emissions Control, as there is no case for any individual approach or any two of the three approaches offering a viable means of the mitigation of AGW.
A starting point of the debate is the need of a stringent Emissions Control treaty, without which, as Wili rightly points out above, the RCP 8.0 massive renewables' deployment doesn't cut emissions since any fossil fuels locally displaced by them would continue to be bought and burnt elsewhere.
Yet Emissions Control alone demonstrably cannot resolve the advanced and accelerating threat we face. As Prof. Mann pointed out on 18/3/14, under BAU we are on track for 2.0C by 2036, or by 2046 if ECS were as low as 2.5. : http://www.scientificamerican.com/article/earth-will-cross-the-climate-danger-threshold-by-2036/
Given that even a cut not in anthro-emissions but in anthro-CO2 stocks will provide a cooling only after the ~35 yr timelag of ocean thermal inertia, even the best case of Emissions Control cannot avoid the hazards of passing the 2.0C threshold.
Taking that best case to be "near-zero global GHG output by 2050" exposes further limitations. Beside the substantial warming unveiled by the closure of of the 'fossil sulphate parasol' the majority of warming from 2050 emissions is not realized until the 2080s, allowing the eight Major Interactive Feedbacks [MIFs]~70 years of continuous anthro-warming, plus warming from their own direct and indirect interactions, escalating their CO2_e outputs far past the point of fully offsetting our best case of emissions control. While the precise levels of those outputs cannot be calculated, the MIF outputs' track records to date strongly affirm this perspective.
For example, consider the paper by Ramanathan et al from Jan 2014 : http://www.pnas.org/content/111/9/3322.abstract
It is a study of the Arctic sea-ice decline fraction of Cryosphere decline in the satellite record since '79, which found that warming from the resulting Albedo Loss was equal on average to 25% of the warming from anthro-CO2 stocks during the period. This is roughly equal to a new China's-worth of annual CO2 output, and given the progress of the sea ice loss it is on a strongly rising trend, but has thus far been masked by ocean thermal inertia. With ~43% of anthro-CO2 outputs going into natural sinks, we should need to cut outputs not by 25% but by 43.8% to offset the warming from ASI decline Albedo Loss.
Given that this is only a part of overall Albedo Loss, and that that is only one of the eight MIFs, it is rather obvious that we are fully committed to the MIF outputs exceeding our best case of Emissions Control if they are not constrained by Geo-E.
Unfortunately even the best case of the simpler mode of Geo-E, "Carbon Recovery for Food Security" cannot provide a timely cooling to halt the MIFs' escalation. While techno options such as DACCS have just their CO2-capture function costed at "at least $600/TCO2" by the American Physical Union ($600Bn /GtCO2), the benign option of native Coppice Afforestation for Biochar offers a second revenue stream via the production of Methanol from the retorts' surplus hydrocarbon gasses.
Yet even with the massive deployment incentives of minimal infrastructure, plus two revenue streams plus raised global food security, and even using efficient village-scale retorts across the 1.6Gha.s of suitable non-farmland identified in the joint WRI-WFN study, there is little prospect of full scale harvests of 10-yr-old growth before the 2040s. This indicates that if this approach was used alongside the best case of Emissions Control, its global cooling could not even begin before the late 2070s, and thus offers no useful control of the MIFs. The real and indispensable functions of a new global industry in Carbon Recovery for Food Security are thus of helping to stabilize global food security ASAP, of minimizing the peak of airborne CO2 ppm potentially before 2050, and of fully cleansing the atmosphere by around 2100, thereby conserving at least a part of the oceans' ecology by limiting the peak and the period of ocean acidification.
The Albedo Restoration mode of Geo-E has had a rising chorus of bad press over the last two years that has been oddly concerted in using standard easily refuted charges - on exactly the model of concerted climate denial. The most obvious of these can be seen in this thread's title, when in reality the very eminent Prof. Holdren (Obama's chief science advisor) while critiquing the sulphate aerosols proposal has pointed out that its costs "could be met by a middle rank economy out of its petty cash" - i.e. there are no vast profits to be made.
While the patently deficient sulphate aerosols proposals are widely cited as the standard option, in reality there are potentially benign options such as 'Cloud Brightening', where seawater is lofted to low clouds as a very fine mist. This option is targettable to provide regional cooling (e.g. of the arctic) and to have any extra rains fall over oceans, and can be halted within the fortnight required for treated clouds to rain out. Its infrastructure costs would comprise around 2,000 wind-powered vessels of 100ft length, and its operation would need to be maintained for the most of the duration of the Carbon Recovery program.
From this perspective Dr Pierrehumbert's shrill denunciation of Geo-E is simply bizarre for a formerly reputable scientist - pretending that we'd be committing future generations to maintaining Albedo Restoration for millennia is intellectually dishonest in assuming, without mentioning, that no Carbon Recovery technique is employed in his scenario.
However, there is a long lead time before any benign Albedo Restoration option could be deployed, consisting of the years needed for negotiating the appropriate governance, for the mandated agency's supervision of proposals' research and their shortlisting for trials, for observations of the chosen option's effects over at least a decade, and then for the negotiation of its full deployment. If that summed to as much as 21 years, it could take us to Prof Mann's 2036 deadline.
The difficulty with this is not only another 21 years of the escalation of the MIFs, whose momentum is then far harder to control (e.g. Methane Hydrates' melt), but that we are liable to see two or more major food producing regions having extreme droughts simultaneously during the 2020s, thus imposing the onset of serial global crop failures and severe geopolitical destabilization. The increasingly erratic annual harvests, going from serious shortages, exceptional prices, national hoarding and unrest to a bumper harvest within a few years indicates that a climatic destabilization of agriculture is already under way and we are already on that track. Whether such geopolitical destabilization generates global conflict is an open question, but it almost certainly curtails the global co-operation essential for the mitigation of AGW, thereby tightening a vicious spiral of hunger and conflict.
From this perspective it is extremely urgent that the negotiation of appropriate UN governance of the R,D&D of Geo-E is formally established without delay. Prof. Forster (an IPCC lead author) wrote an article: http://thebulletin.org/not-enough-time-geoengineering-work7963
that is aptly titled as "Not enough time for geoengineering to work ?" and is well worth reading for a more authoritative background on the issue.