There is still a popular assumption in the Yellowsphere that the world is waiting for XPages and the Domino web server. Like Carl Tyler who think you just have to repackage and rename the software into Websphere something at then start conquering the IT world.
The only success I see is that existing customers will try lowering their license cost.
To attract new customers it would take much more than this.

• Removing ridiculous 16Bit limitations from nsf
• Make nsf significantly faster. There is so much talk about moving everything out of nsf (attachment, images, javascript libraries) because this thing just does not perform
• Much better performance for large data. Being better than Access is not enough
• Performing web pages. It still is hard to find something that is slower than a Domino web server.
• jQuery support. Yes IBM loves Dojo but there is so much jQuery love that it just cannot be ignored.
• It will be hard to win a developers heart if it is not Open Source. Normal people will not get used to nice IBM messages like "The product works as designed" or "There are no plans to address the issue in the current release".
• There are already free as in beer IBM products like Websphere Community Edition, DB2 Express but people who use it normally already use DB2 and/or Websphere.
• There is a common myth in the IBM world that developers love XML files. I have my doubts that this is true
• Much lighter HTML code, a better template engine and reducing the number of layout elements that rely on tables. I really like tables but web developers do not.
• Better integration with other technologies. Authenticating with systems like Twitter (e.g. OAuth), accessing other datastores (JDBC access).
• A broader community that accepts people who do not sleep in IBM pyjamas and is aware that there is a lot of great technology outside of IBM.
• A community website like developerworks/ but not featuring popular forums that have 2 entries in 2011. IBM really screwed Developer works and I am sure they are very proud of it. The reality is that this thing morphed into an IBM product showcase thing where content is not important anymore.
• Trust. Google has a revenue model that is well known and works, Microsoft has one too and although it is hostile in many ways it has always been very attractive for developers. For example you pay for the IDE but not for the code you develop. IBM could have a much better reputation but although they are 80%+ of Eclipse they don't get tired emphasizing how independent the Eclipse Foundation is.
• Young people don't care about IBM anymore. IBM just don't produce technology that is relevant to them. IBM could have bought SUN to change that but we all know that Larry Ellison and Oracle were much more clever (in this case). IBM says they are a mobile player but you won't see an IBM logo on any mobile device (including tablets). Cool Web 2.0 products in most cases can be cool without IBM technology and the cloud is not driven by big servers that were made from IBM. The cloud much is about hardware not coming from IBM. Even worse. Now that IBM is big in quiz show business I believe that young people just don't watch Jeopardy. At least in Germany the show was cancelled when the last TV viewer died of old age.

To add to your last point - I think part of the difference is the change from a top down management style. I've only done some Organizational Behavior study - but the model of a few people at the top making decisions is being replaced by "flatter" organizations.

Younger people want more say, and they want input in major decisions. Because IBM/Lotus are not "consumer" oriented, you see less IBM/Lotus in organizations that have a less top-down management model, and more Apple, MS, & Google. (IE, Universities)

Big companies that have a very top-down governance structure will keep IBM/Lotus, but their users may be frustrated with this fact.

I don't know what is the best course for IBM to take in that respect.

In regards to the other points - I somewhat agree. Personally, if they are not already, I would like to see IBM put as much effort and innovation into each piece of the collaboration stack (especially Notes/Domino) as Google puts into Chrome:

http://googlechromereleases.blogspot.com/

And more importantly, integrate it all into LotusLive, or whatever it becomes. All of it. Full fidelity for Sametime, (and UT), Connections, etc.

One specific thought would be to create a LotusLive appliance that contains a "local replica," to put in offices for increased speed, etc, especially for voip. Have a branch office of 50 people on a dsl connection? Drop in an appliance! (and it would need to be that easy)

--------
These are my hopes and thoughts, not criticism or complaints.