You beg the question: was Soviet agriculture inefficient?

I think the answer to this is both yes and no. The inefficiencies of collectivization seem readily apparent, but it's important to note that some of those were designed inefficiencies. Most of the large state farms focused on staple crops for domestic consumption, sometimes to their own detriment.

The most common statistic out there is that 2% of the land (for private use) produced 26% of all agricultural output. This makes the other 98% of the farm land look really wasteful. But if that 26% is the things that are 'luxury' goods, it starts to look worse. And, there are critics:

http://www.soviet-empire.com/ussr/viewtopic.php?t=47201

The most likely scenario is one of labour focus. The duality of having both private and public lots living in unison. The allure of being able to make profit from your small private plot existing meant a great deal of time was spent cultivating that land because you doing a lot of it by hand, leaving less effort available for your 'day job'. It is possible that eliminating private lots could have improved over-all productivity of the land. On the other hand, the large farms were far less able to service and respond to niche agricultural needs.

However, on the other hand, it's very unlikely, with the control over staple crops being in government hands, that the Soviets would ever face any extreme famines (again), which is an efficiency of another sort.

One factor that sure did not help was that the Soviet government rejected Mendelian genetics and Neodarwinism as "fascist" and "burgeois" "ideologies", and promoted heavily Lysenkoism, a pseudoscienstific form of long discredited Lamarckism. Lysenkoism was the brain child of Ukrainian agronomist Trofim Lysenko, who among other things believed that crops raised in harsh conditions would adapt spontaneously to them and become as productive as those grown in more benign areas. As a result, and despite the absence of positive results, the Soviets poured resources to plant crops in dry and cold areas that returned poor harvests. And they did it for decades. Stalin adored Lysenko and banned anything that contradicted Lysenkoism in academia in 1948. No Soviet scientist dared to criticize him openly until Khruschev (also a supporter of Lysenko) fell in 1964. You can imagine what that did to a generation worth of development of agricultural sciences in the Soviet Union, considered that Lysenko himself was the director of the Academy of Agricultural Sciences and the Institute of Genetics for long periods between the 1930s and the 1960s.

It's easy to see why Lysenkoism appealed to the Soviets on an ideological level. Besides nationalism (Lysenko was Soviet, as was his teacher, Michurin), under Lysenkoism all plants can be raised to become productive members of the state, not just the ones whose parents were better at the job. Just like people!

S. J. Gould published an essay that went deeper in Lysenkoism and several possible factors that influenced Michurin (for example, that working in vastly spaced Russia made it harder for him to appreciate and value the competence between individuals as a leading factor in natural selection).

Ludwig von Mises wrote an excellent book in 1920 called Economic Calculation in the Socialist Commonwealth, in which he argues that the key flaw to central planning is that market prices cannot act as information signals to optimise input and output.

In other words, without a market denominated price, the aggregate information that millions of market buyers and sellers each have regarding demand and supply is not used to determine the economic output, which is instead directed by a small committee with much more limited information.

This leads inevitably to gross inefficiencies, oversupply in some areas, undersupply in other areas, all aspects that are not compensated by increases and decreases in price until it's far too late in the day to do anything about them.

He then went through a series of possible scenarios about the political outcome of such errors, most of which led him to conclude the system would become more and more rigid until its economic collapse, which he estimated to be about 50 years. He wasn't far off: the system lasted 70 years. The difference is relatively easily attributable to the fact that many Soviet economic analysts looked at market prices in the West to give them an idea of possible errors in their planning systems.