What are these "significant drawbacks?" I can't think of any that are significant.
Volatility. Gaining and losing tens of percent in a day is great for speculators, but it's terrible for people who simply want to earn money and spend it on goods. There are several possible ways to address this (eg, EnCoin, GEM, and my own proposals). None are tested, but if they work out they could be very successful. As a related benefit, they would eliminate the cries of "pyramid scheme" which have made Bitcoin a running joke among the people who should be it's strongest supporters.
The volatility is purely market based - and this means the volatility is valid price discovery. If you think the price at any given point is flawed, you can take an opposing position. Bitcoin's volatility is not a flaw of Bitcoin, by any measure. In fact, it's fluidity and ability to be traded by anyone, anywhere, quickly, is one of its shining attributes. Any "means" I can think of to "curb volatility" would be harmful, because they necessarily involve the prevention or reduction in the ability of someone to take large speculative positions. Speculation, contrary to popular belief, is very healthy and the fact that anyone can speculate in Bitcoin is not a bug, but a feature. And of course, volatility will recede with time necessarily, so even if it is a problem, it's one that fixes itself. At most, it is a temporary drawback, not a "significant drawback."
Mining is very expensive, and all BTC holders pay for it - currently about $36,000 per day - as inflation / devaluation. The effect is currently overwhelmed by speculation, but I don't think the speculators will pay for it forever. Increasing exchange rates make this worse. One possible alternative keeping with an overall Bitcoin-style network is proof of stake instead of proof of work. Other more radical systems are possible as well.
Not true. Your $36,000 figure assumes everyone is paying the average kwh $ for electricity. The Bitcoin system encourages mining by those with the lowest marginal cost of electricity, and thus the real "cost" of mining is
far lower than your estimate. And the inflation issue is one which, again, fades with time. The inflation goes away. I see no significant drawback in the mining system or inflation schedule.
Anonymity. It's good enough to protect thieves who steal large amounts and are willing to spend time and money laundering their coins, but at the same time it's hard for the average user who simply wants privacy to get strong protection without effort and expense. It's a poor balance; going in either direction would help.
Bitcoin is the most anonymous way to buy something online. How is "anonymity" a "significant drawback" of Bitcoin, when it's the best available option? The average user doesn't need "strong" protection, he just doesn't want CamSluts.com appearing on his credit card bill, and Bitcoin is perfect for that. Users who want serious anonymity can obtain it fairly easily - but let's remember anonymity is not a cheap commodity and there is a cost to acquire it. I think Bitcoin has brought that cost down substantially.
Long-term viability. It's not certain that transaction fees will be an adequate way to protect the network in the long-future. Right now 51% attacks are very expensive relative to the size of the money supply. When generation goes down significantly, we may have a problem. It may be necessary to apply a new policy such as: high fees (limiting the utility as a currency); or reinstating generation / inflation (thus screwing all the people who are speculating that the limited supply will make them rich). While double spend attacks can probably be defended at a reasonable price, it may be economically viable for a large entity with an interest in destroying Bitcoin (central banks, large commercial banks) to DOS the network by buying hash power.
The fees, long term, will be priced by the market, meaning they'll be the minimum miners require to mine. It's "possible" that the long-term market price of mining is so low that a 51% attack becomes dangerous, but suggesting it now is just speculation, and cannot be said to be a "significant drawback" when there's no solid evidence that it will ever occur. It's a "possible drawback", true.
Security. Take a look at how hard it is to get something like P2SH accepted (which is looking increasingly likely but it's still not a done deal) - and despite people's hopes, it's not a silver bullet. It will help people with very large accounts, but I expect it to be prohibitively difficult for the average user to use.
The new Armory client already has multisig, offline wallets, easy paper backups, etc. The security of Bitcoin seems pretty darn good thus far. Thefts are news, meaning they're rare, and they always deal with some poor security that is not the fault of Bitcoin (bad VPS company, for example). Again, no "significant flaw" here.
... Just to name a few off the top of my head. None of those are deal killers, but they're all significant drawbacks. Making the fundamental changes to correct them may be too controversial to accomplish in Bitcoin, and an alternative coin may have to do it instead.
In Bitcoin's favor, it has jumped out well into the lead of cryptocurrencies. I find the current market adoption disappointing (which is what I base my low valuations on), but there is significant infrastructure built up, good name recognition, and an enthusiastic core of community support. That's a lot for a newcomer to overcome, but in a global market sense Bitcoin is still miniscule, so don't mistake it for an invincible force.
Bitcoin is by no means invincible, and it's future is highly speculative, for sure. But I still see no significant drawback to it's structure. I see a handful of small problems, all of which have pretty good solutions. For a long time people said the "waiting for six confirmations" was a significant drawback, but most people understand that problem has been almost entirely eradicated. There's an important lesson there.
My view is simple: the likelihood of a wholly new coin being vastly better than Bitcoin is less than the likelihood of Bitcoin overcoming those few issues it struggles with. Thus, while Bitcoin may certainly be destroyed by a competitor some day, I see no evidence at all that such a competitor is anything more than theoretical. And I think the myriad problems Bitcoin has solved is too often taken for granted, and too often overlooked by those who focus instead on the limited number of problems not yet overcome.