However, I don't think there is anything wrong with leaning your choices one way or the other. Everyone sets up a diversification allocation based on their risk comfort level. It makes sense that your comfort level changes depending on whether the market is very high or very low. Changing the allocations a little (but not a lot, for all the reasons others have outlined) to me makes sense. It may help your portfolio a little, with a relatively small risk of losing some gains compared to a pure "no-timing" strategy.
That sounds very good but the problem is knowing when "the market is very high or very low." It seems like it is possible but it is only with some perspective that the idea of high or low makes sense.
Is it high or low right now? We may know in 5 years or so.
Since the late 80's the Nikkei has been basically flat but has had some large moves up and down. Do you really think if you lived in Japan and primarily invested in the Nikkei, you would have been able to predict which move was a breakout up or down - or was nothing more than noise in a long sideways move? The reason many people believe they can do it in the US is that the long term trend of the market always bails them out. No one knew in 2000 or 2008 how low the market would go or how long it would stay down.
On the other hand, my Dad said that when the Dow crossed 1,000 for the first time, everyone thought it was insanely overvalued and maybe it was but I would ask if it would have been a good idea to shade your allocation down? I would ask the same question at a Dow of 5,000, 10,000, and 15,000.
That being said, I opened a 529 plan in 2009 with the maximum contribution. It has more than doubled but I don't kid myself I had correctly valued the market. It was pure luck. I also try to add to my equity holdings when the market backs off significantly but I never know if it will work out. One of these days we may head into a Japanese style sideways market that lasts for decades. If that happens, I bet there will be a distinct lack of people claiming they can tell when the market is under or over valued.