There are a bunch of things that I’d like to see Embarcadero do. I’ve listed and discussed some of them below. I’ll probably think of more later. 
They are in no particular order, and they are not grouped in any particular way. They are a bit random, and range from business decisions to minor technological decisions. Where it makes sense, I’ve linked the titles to the entries on http://delphi.uservoice.com/ so that you can vote for the items if you see fit.
I’d like to see Embarcadero:
- Provide an Enterprise-level MVC web framework for Delphi. This almost seems like a no brainer to me. Ruby on Rails has had a profound impact on web development and development in general. In the .Net world, MVC has become the leading ASP.NET development method,winning over hearts and minds from WinForms. The general idea of MVC is becoming the norm for much of the development world – separate those concerns!. Delphi’s new RTTI capabilities would actually make this kind of framework very, very possible and very, very cool. And Delphi already has a very powerful and capable web infrastructure to build on: good, old, and venerable WebBroker. There are even existing frameworks out there that could be leveraged, including the G Framework. There is a business opportunity here for Embarcadero, if not for an enterprising third-party Delphi developer. A native, ISAPI based MVC framework in Delphi? That would be very, very sweet.
- Stop trying to do other stuff and invest in Delphi. This one has been one of my huge frustrations over the years. Delphi is a profitable product, but no matter who owns or runs the show, the profits always seem to get skimmed off to fund other “next big thing” projects of, well, questionable benefit at best. Remember “SDO” taking the market by storm? Yeah, me neither. AppWave seems to be interesting and have a lot going for it, but I personally would rather have seen that effort invested back into Delphi. I hate to think about where Delphi might be today if it hadn’t been used to fund other projects and instead been allowed to use it’s profits for it’s own development. Delphi could use a year or two of un-distracted attention from its owner.
- Create a Javascript/HTML development tool. I’ve always said “As assembler is to the Intel chip, so Javascript is to the browser”. James Governor has it right: “Learning Javascript used to mean you weren't a "serious software developer". Today, not learning Javascript means the same thing.” A powerful, feature rich RAD development tool for Javascript and JQuery would be really cool and a great new product for Embarcadero. Maybe RADPHP could be steered in that direction?
- Make Dependency Injection part of the RTL: If you haven’t figured it out yet, I’m a big believer in Dependency Injection. So much so, that I think that if you aren’t doing Dependency Injection, then you are doing it wrong. Incorporating a DI Container into the Delphi RTL would be very cool. Again, Generics, anonymous methods, and the new RTTI makes is very easy and powerful, and there are open source projects to leverage and build on (like my favorite, Delphi Spring).
- Do one thing at a time and do it really well: This is the one that I wish they would do the most: Take the product forward one step at a time. For instance, the next release should have as it’s main focus 64-bit Delphi and only 64-bit Delphi. That’s it. Other improvements can be made, of course, but clearly that should be the “big one”, and it should be the only “big one”. Don’t try to do two or three “big ones” in a single release. Make each release focused on a big, single step forward, executed thoroughly and solidly. The product will be fine as long as it shows steady, sustained improvement. Many large improvements executed all at once is not what the market wants. Focus and deliberately move forward.
That’s all for now – I’ll probably have more as I think of them, but that ought to be food for thought for while