The success or failure of the Rooftop solar systems

There are couple of issues that are emerging:

Rooftop solar systems, indeed, are best solutions to thwart the power crisis. Idle roof is used and hence no problem of land issues, a household can be power-independent, etc
Subsidy menace is doing more damage to solar technology than doing good! The strength of solar technology is its weakness as well. Unlike conventional power plants, solar energy can be generated anywhere. It's modular. The solar power plant can be as small as 1 KW and as big as 100MW and more! Conventional power plants take couple of years to build (3-4 years) but solar power plants of megawatt size could be built in just 6 months max.
As solar systems have been Govt affair' for long (three decades!) hence quality and performance has been an issue for such a prolonged time. Private solar companies have not been putting their heads together to bring in solar product/system standardization (I agree with Suresh Tripati)... Solar module technology is robust, but not the EPC as such. JNNSM has left some thorns in the 'flesh' of solar industry!
We are in the transition as far as the solar technology infusion into market is concerned. There are issues such as AC vs DC and the related wiring problems in the households. We should certainly address the packaging of rooftop.
Solar rooftop in conjunction with LED efficient lighting solutions (as pointed out by Mehta), DC fans etc have to be brought into the foray of total solution.
Though grid interactive systems (with net metering) could solve the major problems, unreliable grids do not auger well for the synchronization. Knowing this, if we go for battery back up, it's messy and we 'import' problems associated with battery maintenance.
It's sad that the fossil fuels are dwindling. Solar cannot be supplementary energy; it can at best now become complimentary energy in the energy mix. So, time is running out with the decreasing coal stock and drying river beds. Without conventional energy flowing in the so-called power lines, solar cannot be synchronized. This means that we should speed up solar power generation (off-grid and on-grid) to a maximum extent before fossil fuel resources totally dry up.
We are on tech cross-roads as far as solar is concerned. But yes, the great time has come for solar tech with enormous awareness of it in the public domain. Thanks to JNNSM and frequent backouts and brownouts.
With excellent packaged solar rooftop solutions, there are multi-fold benefits; power crisis, to a greater extent, can be mitigated (of course, with CO2 mitigation) to a large extent, solar industry grows.

All said and done, the solar actors and the players have to live up to another challenge: The conversion efficiency of solar modules and inverters has been stagnant for, at least, two decades. We have to pull up our sleeves to increase the power generation from 1.6 million units/MW to, at least, 2 million units/MW whether at the module level or in the system integration level (by increasing the system efficiency through maximum power transfer theorem and so on. I believe it's possible to work on capitalizing on reducing the impedances at various levels).

We cannot keep talking only about the age-old technologies and the subsidies to 'push' the technology. Many conferences talk about how to harness the subsidies and incentives, and how to carry out EPCs. But none have framed to talk about breakthrough technologies. There must be rapid changes in the approaches; technology-driven and innovation-driven methodologies are the answers. India has to drive on these paths, if it's solar dream has to come true to address the energy issues.