How about if I explain my position more completely, then you can actually explain yours instead of merely stating it.
First and foremost, when talking about blockland we need to bear in mind that it's not so much a "game" like basketball as it is a "toy" like block-Os. This shifts the desirability away from strategic decisions and toward the creative expression of the person using the toy. The issue then is not, say, wether or not floating bricks would imbalance the "game", but rather wether or not floating bricks are or can ever be part of a desirable creative expression.
When examining any sort of online entertainment, we see a large tendency toward customization of the play experience. In Source engine games, plug-ins not specifically associated with the game in question are often installed on host machines to allow administrative tasks not normally available according to the standard game, as well as extra UI features. In online strategy games, one can host a game with a great myriad of options to the game in question, such as random/fixed start positions, starting resources, fog of war behavior, etc. These options are not ignored; people use them to the fullest. In BL itself people shove things into their "Add-Ons" folder for basketballs and portal guns and hookshots. Clearly there is no one "perfect" game, and everyone has their own preferences.
The question then becomes when additional preference becomes a bad thing. Nothing created in blockland can't be created in a 3d modelling program, but it's much more time-consuming and takes more knowledge of a given program; this is enough for most people to think of as "not fun". So, by limiting player actions, we can make things simpler and easier.
Another issue is consistency. Even if all possible option sets for an online game are simple, if they vary widely enough and are not clearly different at first glance, people can become confused and upset when the gamespace doesn't behave how they expect.
The first issue, simplicity, can be dismissed easily enough; Everyone can understand floating bricks. People opposed to floating bricks as an option habitually fail to explain their reasoning, but occasionally I can get an inkling toward issue number 2, consistency (in the form of, for example, "it would be too hard to judge a screenshot for quality"). To this I say, nonsense. If it's that big an issue, it can be a Yes/No flag right next to Ded and Pass on the server list. We can't prove that a good solution to immediately-recognizeable floating in one form or another
can't be done - only that we fail to think of a way - and therefore the idea should be rejected upon such failure, rather than outright.
edit: I apologize for straw-manning the opposition, but they've yet to give any rationale whatsoever, so I have no choice but to speculate on possible arguments.
for anyone who wants to post reasons (
reasons) why floating bricks should not even be an option for servers.